Now fast forward to 2026; “Intelligence” is instantly associated with specific AI technologies (usually either LLMs or Agentic AI right now), yet all of the key Healthcare challenges listed for 2010 still exist to some extent, especially the last one. In other words, while we’ve been solving at least some of the technical challenges (mostly with legacy technologies but some now with AI), we’ve not made any real progress towards making Healthcare more effective than it was before. In some cases, Healthcare in the United States may have actually gotten worse since 2010 given the shortages of doctors and closures of 100’s healthcare facilities across the country. And beyond those structural problems, the core metrics for Healthcare outcomes in the US have not improved either (Life Expectancy has actually dropped, Infant Mortality keeps rising and most other metrics still lag significantly behind peer nations, except for medical mistakes). And all of these relatively dismal metrics come despite the fact that the US is spending twice as much per capita on Healthcare as most of those peer nations). So, 2026 seems a lot like 2010 and 2010 seemed a lot like the 1990’s and so forth. In other words, technology isn’t making Healthcare more Intelligent, yet.

Why has the US Been Stuck in this Situation for Decades; What’s Wrong with Our Healthcare System?

If we ever wish to make Healthcare truly Intelligent (and thus effective across any set of objective metrics that we might wish to apply to it), we need to understand what works here in the US and what doesn’t. Only then can we begin to determine how or even whether technology can solve some of those problems. So, here are the big (overarching) problems with US Healthcare as I see them:

1. There are few if any incentives tied to outcomes.

2. Access to Healthcare is a still a problem, but it’s not the main one.

3. Some aspects of how we manage Medical education are hopelessly out of date.

4. We still collectively don’t understand how to deal with complex patient scenarios.

5. We’re not willing to address many of the reasons why costs are actually exploding (while outcomes remain poor or are even getting worse).

When I refer to Healthcare it’s worthwhile to step back for a moment and define what that represents (both here and elsewhere):

Healthcare - The combined institutions, practices and associated capabilities available to manage the health of people living within a society. This includes both Public Health organizations and the Private Health marketplace and all subdomains within both of those (e.g. Providers, Payers and Oversight). We refer to this as Healthcare and not Medicine as Medicine represents a practice with the larger domain of Healthcare.

The first realization when we look at some of the core problems as well as the definition for Healthcare is that much of this topic may extend well beyond the scope of technology. In other words, while technologies are used in every facet of what’s been described, it’s not clear that every problem associated with Healthcare can (or should) be resolved by technology in isolation. Let’s look at “Access” for example; in this context, Access refers to the ability of people to obtain healthcare services. The obvious problem with this definition in terms of how it’s been viewed in the United States is that Healthcare Access has historically been conflated with Access to Healthcare Insurance. While the two are related, they aren’t quite the same. In other words, Insurance doesn’t guarantee access to all types of care or perhaps more accurately any care that falls outside a person’s ability to pay. The reason why this distinction matters is that many people felt that major Healthcare Reform had been achieved simply through the expansion of Health Insurance Access (in the A.C.A or Obamacare or even Medicaid). But that’s not really what happened; while there was an expectation that overall Healthcare outcomes would improve with greater access and costs would also generally decline - those things haven’t happened because the ACA (and by extension additional coverage through Medicaid) was only looking at the problem from one perspective. Any meaningful Healthcare Reform necessarily must look at the holistic landscape, rather than one portion of it if we want to improve outcomes and make US healthcare as effective and cost efficient as that provided by our international peers.

Getting back to the main question of this section - the reason why the US has been stuck for so long with a Healthcare system that despite tremendous advances in some respects, seems to be just as inefficient as ever, is that the problems associated with it span the entire ecosystem and both the Healthcare community and political leadership have been unwilling or unable to view the dysfunction in its proper (holistic) context. When viewed through the lens of what technology can and can’t do to help solve those challenges, the question then becomes more complex and potentially equates to a Catch-22 situation (if the expectation is that technology-driven innovation on its own can correct systemic problems without actually addressing the root causes of those problems).

So, What Constitutes Intelligent Healthcare in 2026?

Is it the ability to apply various types of AI technologies to resolve long standing problems? Yes and and no - but Intelligence in our current context places the Healthcare Domain first - in other words, it’s Intelligence about how we address the problem space rather than viewing it as Artificially Intelligent Technology (which may or more not resolve what resides in that space). This may seem to be a bit nit-picky, but the making the distinction matters because it helps us to calibrate both our expectations and focus in relation to how to get to a better Healthcare system. So, in some respects, Intelligent Healthcare in 2026 isn’t that much different from how we began viewing it in the 2010’s - but there are some interesting new opportunities associated with it.

Put another way, the starting point for viewing, managing and improving Healthcare systems and outputs is the willingness to view it as a complex, system of systems. Any attempts at individual improvements or reforms need to take into consideration how that focal point or component interacts within the rest of the ecosystem. If we wanted to catalog what would make Healthcare Intelligent, it might then include the following characteristics:

1. It’s holistic - both from a domain perspective and from a patient perspective.

2. It’s focused on continuous improvement (in terms of patient and system outcomes).

3. It’s evidence-based (in a way that current medicine isn’t; we’ll come back to this later).

4. It’s pragmatic (rather than idiosyncratic).

5. It’s both technology agnostic and innovative at the same time.

We’ll look at each of these in turn in this and several follow up articles.

Holistic Healthcare - When most people hear something like this they might think about Holistic or Alternative Medicine, which of course might fit here as well, but what we’re really referring to with this term is being able to view the entire domain of Healthcare as a connected set of lifecycles. Perhaps the best way to illustrate this is through an example in relation managing infectious diseases…

• Public Health agencies, (there are nearly 5,000 of these in the US alone), help to identify and mitigate specific disease outbreaks (using both guidance and action).

• Healthcare Providers and Members of the public help to avoid or contain outbreaks through precautions and preventative health protocols.

• Medical professionals then treat cases as they arise managing patients and facilities to ensure that the appropriate care is available throughout the duration of outbreaks.

Each of these steps is equally important and just as important is the ability to recognize that this is just one of many such ‘lifecycles’ wherein all of these participants engage in related Healthcare activities - all of which contribute to overall national Healthcare (system) outcomes. Each of the previously described characteristics also includes or offers opportunities for technology to improve current practice. Thus Healthcare doesn’t have to be “Artificially Intelligent” to qualify as Intelligent, but can still potentially utilize AI.

What Can We Do With AI Now That We Couldn’t Do Before?

First, it’s worth noting that we could have done much more before even with the legacy technologies had we been willing to view the Healthcare domain differently. It’s also important to note here that applying Intelligent principles to Healthcare doesn’t or shouldn’t occur in just one direction either; Intelligent Healthcare is both a bottom-up and top-down process. In other words, while national policy is important, it can’t be the sole focus of reform if we expect that reform to succeed. Having said that, it’s also important to acknowledge that there are long-standing challenges that can be addressed now that were out of reach before as a result of the introduction of new technology (including AI). Before we start looking at that though, it’s important to check our expectations - especially when it comes to Healthcare. The AI industry is now in the midst of no little amount of controversy due to capability claims leap-frogging actual performance, so when examining how and where this type of new tech can be applied to Healthcare, it’s crucial to stay grounded in the near-term reality. Yes, maybe someday we’ll get to AGI and all of the things associated with that - but we’re nowhere near that point yet (despite the hype).

I gave one example of how this might play out (more from a bottom-up perspective) in our sister publication; The IT Architecture Journal. In that series of articles, I looked at how AI-related adjustments to Healthcare portal and library-focused solutions might significantly improve the ability to discover new insights from existing repositories of Medical information. That targeted Use Case roughly aligns with being pragmatic and holistic. In other words, being able to link together information and evidence in new ways with less effort (in terms of applied integration) has tremendous potential - and it doesn’t require a ‘thinking’ replacement for doctors, researchers or anyone for that matter. It’s an example of an Augmentative use of AI. This is precisely where most of our near-term opportunities lie.

So, what can we be working on near-term with newly available technology? Here’s a short list of generalized Use Cases:

• Improvements in Healthcare integration and interoperability.

• Improvements in Healthcare Collaboration (bringing Healthcare communities and professionals together in new ways, dependent on the previous point). This is especially important when viewed from Patient and Public perspectives.

• Improvements in Care Outcomes if we decide to bring in more “Evidence” and use that at a co-equal status with study data.

• Improvements in Healthcare Cost Efficiency - if we allow data insights to drive policy decisions (e.g. not devolving all such discussions into mere political squabbles).

• Improvements related to Education and Preventative Healthcare (e.g. “un-priming the pump” to ensure that primary care doesn’t always end up occurring in emergency contexts).

• And opportunities to change / evolve out of various Healthcare related idiosyncrasies by helping to shine a spotlight on them.

What we absolutely don’t want to have happen is what’s occurring now in other industries where organizations are using AI as an excuse to eliminate mission critical personnel under the mistaken belief that AI can do their jobs (because in most cases, they can’t). Healthcare is already dealing with major shortages of key personnel and we don’t want to exacerbate those existing crises.

In several follow-up articles, I’m going to walk through the General Use Cases outlined above in more detail.

Copyright 2026, Digital Perspectives

Stepping back to an even larger context - a $55 billion military spend (actually $74 billion) makes this 2027 budget line item for drones larger than the 8th largest defense budget in the world (France’s entire military). Sure, we were behind somewhat on drones over the past few years, but is this colossal spending request actually necessary? This week, we’ve been looking at several related military topics with an article on the Golden Dome here and a previous one on US Military Doctrine in our sister publication - Political Perspectives.

So, Why Not Spend a Lot on Drones?

Good question; it’s clear that we had been somewhat behind not only in procurement but in adoption of this technology which is rapidly and radically reshaping modern warfare before our very eyes in real time. But when considering that is a 24,000% increase in just one year, it’s valid to ask whether there is any serious planning behind this request. All preliminary indications are that this spending increase seems to be an arbitrary number with few if any details available as to where it would go. Here’s the DoD press release description on the $445 billion budget increase for 2027 (and the other line items are similarly sparse):

Drone Dominance

Delivering over a $74B investment, President Trump’s request triples how much we will spend on drone and counter-drone technologies from FY26. This marks the largest investment ever in this sector.

Ok. It’s worth noting here that the $55 billion from the press releases (for something called DAWG) is only part of the overall proposed Drone spend which we can see here is actually $74 billion for FY 2027.

Here’s some more detail on DAWG (which as far as acronyms go, is one of the worst I’ve ever seen for the DoD to date):

The Defense Autonomous Warfare Group (DAWG) is a specialized Pentagon organization, established to succeed the Replicator Initiative, focused on rapidly acquiring and fielding thousands of autonomous, AI-enabled, and unmanned drone systems. Led by Lt. Gen. Francis L. Donovan, DAWG aims to integrate these technologies into military operations, particularly for Pacific threats, supported by a significant FY2027 budget request of $54.6 billion.

So, there’s a couple big things wrong here that stand out up front, including but not limited to:

1. The fact that the DoD is investing $55 more dollars into something that they haven’t yet properly defined (e.g. Autonomous warfare). We talked about this in in a prior post here on Digital Perspectives.

2. Going from $225 million budget for a single program to $55 billion in just one year when the true mission hasn’t even been fully defined or fleshed out yet, is well - a bit wacky and unlikely to do anything other than waste a great deal of money really, really fast. In the DoD, money has to be spent in the year its allocated for and if it isn’t - it goes back to the Treasury. This situation provides (and has always provided) a huge incentive for DoD programs to spend everything no matter what is being done with it before they reach those deadlines. It’s a golden invitation for fraud, waste and abuse.

3. We’re probably going in the wrong direction in relation to drone application within military doctrine right now. The US has not properly absorbed the lessons of the Ukraine War yet (and this has become glaringly obvious in the Iran Conflict thusfar) and until we do, we may end up creating new liabilities rather than innovative solutions.

This last point is the most complicated of the three and merits further explanation…

US Drone Doctrine is Currently Confused

The biggest drone-related lessons from the Ukraine War seem to be lost on our military so far. Yes, the Ukrainians (and Russians to a lesser extent) have deployed a variety of new autonomous and semi-autonomous drone systems successfully - and the pace of these innovations keeps accelerating. However, the real revolution in drone warfare is not autonomy - it’s asymmetry. While the DAWG and its predecessor the Replicator Initiative are talking about deploying thousands of highly sophisticated drones - Ukraine is deploying 7 million far simpler drones this year for a fraction of what we’re about to spend. It’s worth noting that at least the Replicator program seemed to understand the need to focus on low cost solutions - that part got tossed with the introduction of DAWG. And keep mind what happened in Ukraine since 2022 - in the first year, all of the drone focus was on high-end, expensive drone systems - most notably Turkey’s Bayraktar drones. Those drones cost up to $7 million each and Ukraine was only able to purchase a limited number, so Ukraine quickly developed their own drone industry and focused on far less expensive models. The US is not following that lesson learned.

This value proposition is a pretty huge point to miss - but somehow the US is missing it now. Trying to turn drones into pseudo fighter planes or flying AGIs or both is a great way to pay homage to the Terminator franchise, but it’s not going to return much value to US warfighting capabilities (in comparison to the massive investments) - especially when our adversaries choose to pick the pragmatic path and we don’t. But there are other points just as important here to consider and of those the most critical one is Drone Combined Arms Integration. The other revolution that Ukraine has introduced has been in the myriad of ways that they’ve adapted various drone technologies to support nearly every aspect of the war (and how it’s being coordinated not merely at the tactical level but across all strategy now as well). For Ukraine, these innovations have taken about 4 years to invent and successfully implement. The US does have the benefit of watching all of this happen from afar, and can even hire Ukrainians to help us - but at some point there needs to be an internal effort here to try to integrate those approaches with what we’re already doing. That hasn’t happened yet and it seems as though DAWG might be going in the wrong direction.

In other words, initial indications are that DoD leadership do indeed want to recreate Skynet by deploying a special command of just uncrewed, autonomous systems (making all of our dystopias a reality). But this sounds a lot like we’re creating a new branch of warfighting as opposed to fully integrating it into our existing structures (and then helping those to evolve properly). Drones are in reality ubiquitous in the way that computing and thus Cyber Security have become ubiquitous. Yes, it’s important to jump-start production and facilitate innovation, but that must all be done with an understanding of the best Use Cases for that technology (as the associated integration must occur to fully realize the intended benefits).

What Should We Be Doing Instead?

Instead of trying to reinvent a wheel that’s just been invested (in Ukraine), we should pivot to their models first; both for rapid production and for Combined Arms Integration. We could do both in 2027 for a fraction of the proposed $55 billion spend on DAWG (probably for around 1/4th the money). This would also give us time to work through issues on organizational adoption (and structural evolution), autonomous doctrine as well as further US-specific innovations. Let’s not throw away another $40 billion or more that we don’t need to spend, (Golden Dome is bad enough). Let’s do drones right.

Copyright Digital Perspectives

The History: Reagan’s Star Wars Program Evolving into Missile Defense

1983 - That’s 43 years ago - that’s when this Science Fiction idea was first announced and coined “Star Wars.” The official title for the program was the Strategic Defense Initiative (SDI). BTW - the entire DoD budget for FY 1983 was $258 billion (that translates to $855 billion in today’s dollars, so the real DoD budget has now risen to levels higher than those under Reagan as he outspent the Soviet Union into oblivion).

The Strategic Defense Initiative (SDI), announced by President Ronald Reagan on March 23, 1983, was a proposed U.S. space-based missile defense system intended to protect against nuclear attacks, aiming to render ballistic missiles "impotent and obsolete". Nicknamed "Star Wars" by critics, the research program explored futuristic technologies like lasers to replace the doctrine of Mutually Assured Destruction (MAD).

It’s worth noting that estimates range between $60 and 260 billion (in 1980’s and 1990’s dollars) for the total cost of the SDI program, but no one knows for sure and that figure only covers the initial program (and not any related follow-on efforts). We need to add at least another $200 billion for the Missile Defense Agency (MDA), which inherited the mission and technology from SDI sometime around 25 years ago. Also worth noting, the original premise of using space-based lasers to shoot down ballistic missiles, as described by Reagan in his speech was never achieved in any form. So, the entire $500 or so billion spent to date on these programs has been dedicated to missile interceptor-related technology. And one last note here, none of these of programs had any impact whatsoever on Mutually Assured Destruction (MAD) as the US, Russia and China have just as many nukes pointed at each other as ever (with no effective way to stop the vast majority of them from reaching their targets).

SDI was a massive failure and the program was terminated in 1993 - achieving none of its stated objectives (except perhaps in contributing to the Soviet Union’s financial collapse as they tried to match our spending here too). MDA hasn’t done much better; although its mission to manage short-range inceptors (like Patriot missiles) has been at least partially successful. This short and medium range missile defense is now often referred to as an Iron Dome as that’s the name given to it by Israel. We’ll come back to the Iron Dome in a moment.

But let’s go back to 1983 first. Why did anyone ever think that space-based laser technology as described by Reagan was possibly viable within a reasonable timeframe? It obviously wasn’t and still isn’t viable now - 43 years later. But SDI was one of the first examples of the Fanboy Effect; where the perception of reality is skewed by popular Science Fiction; we see lasers in movies and TV ever day blowing up Death Stars, so it must be possible, right? The contractors and DoD weren’t going to argue with the President and tell him he was wrong - especially when there was so much money to be made, yet it’s pretty clear that the scientific community knew it wasn’t going to work from the get go.

Another important consideration here (that remarkably is seldom addressed) is how well this had to work to be truly viable. In other words, for a missile defense shield to be effective against nuclear attack, it needs to stop 99.9% or so of all incoming missiles. Let’s do some basic math; if our enemies launch 10,000 missiles (or perhaps fewer but many with multiple warheads or MIRVs), and we are able to destroy 99% of these - about 100 thermo-nuclear warheads still end up reaching their targets - destroying ever major city in the US and sending us back into the Stone Age. So, 99% effectiveness just doesn’t cut it, we’d need to be around 99.9% or ever better to achieve a functional nuclear shield - that’s a pretty high accuracy / effectiveness goal in most systems, let alone the type of system we’re describing here.

What About the Iron Dome?

People have pointed to Israel’s missile defense shield as the model for the Golden Dome for a variety of reasons, but unfortunately none of those reasons makes any sense; here are a few of them:

1. It was never really Israel’s shield, it was a variation of our MDA and was totally dependent on the short range interceptors now being managed by the MDA. These missile on missile inceptors are good, but come nowhere near 99% accuracy (and that accuracy degrades as munitions run out and / or enemy munitions overwhelm them). In other words Israel didn’t invent some new magically effective solution at all and we’re not trying to emulating them, because they’re already emulating us.

2. The Iron Dome is relatively ineffective against high-speed ballistic missiles - and if any of those were nukes, Israel would have been annihilated a long time ago. In other words, missile on missile interceptors can never achieve anything close to the 99.9% to make a nuclear shield viable.

3. And over time, as the Iron Dome has evolved, so too has the race to successfully evade it. Anything short of the magical laser-based solutions that we’ve been chasing for four decades is doomed to degrade over time. It’s also possible that this could also happen to a laser based system too if one is ever successfully deployed - but that just hasn’t happened yet.

The Iron Dome was a mixed-up metaphor that led to our current Golden Dome proposal. And even as the proposal played, we’ve witnessed the Iron Dome become less effective in real time as the current Iran conflict plays out. BTW - what is that proposal anyway?

The Golden Dome

Despite the idiotic rebranding by King Midas, the Golden Dome is essentially the confused attempt to combine current mid-range defense (MDA & Iron Dome) with the long-term space-based dream of SDI. Depending upon which description you look at, space-based high-energy lasers are sometimes referenced, but perhaps it’s not overly hyped now because defense contractors are keen not to remind folks of the original failure of SDI.

The Golden Dome is a planned, multi-layered US missile defense shield initiative introduced by President Donald Trump in 2025 to protect the homeland from advanced threats. Modeled after Israel’s Iron Dome, this "system of systems" combines terrestrial interceptors with space-based sensors to destroy ballistic, hypersonic, and drone attacks. It involves a four-layer approach, integrating land-based radar/interceptors, sea-based systems, and space-based interceptors aimed at destroying missiles during their boost phase.

And how much is all of this supposed to cost? Well, nobody seems to know. Estimates have ranged from $185 billion to $3.6 or 7 trillion to whatever it takes. The Congressional Budget Office (CBO) now places total cost around $542 billion, yet the FY 2026 budget only included $17.1 billion for the Golden Dome program (but that’s separate from the MDA’s $13.2 billion FY 2026 budget, so $30 billion overall this year just for missile defense). The scope of what’s needed to deploy such as shield in space still isn’t clear either, with some estimating 7,000 launches required and thousands of additional satellites in orbit (on top of the thousands that Musk, Bezos and the Chinese are sending up anyway). And will it work? Nope - let’s examine why.

Why The Golden Dome is Doomed

There are a number of reasons, but we’ll focus on the big ones here:

1. Much of the key technology (to get us to 99.9%) simply doesn’t exist. We can say AI will solve it all for us, but we were saying computers would solve it all back in the 1980s. Wishful thinking does not count as engineering.

2. Even if we solve the immediate problems associated with a space-based system, our adversaries will find new ways to defeat the shield and most likely at a much lower cost than the shield itself. This would make our Golden Dome like an orbiting Maginot Line. It’s worth remembering here that many previous innovations were designed specifically to thwart missile defense; things like MIRVs, Hypersonic Missiles and drones (and not to mention EW counter measures).

3. The massive complexity and cost of the program will ensure that it falls behind; behind both our need for it and our adversaries’ abilities to counter it.

4. Deployment of this in any form will also lead us to a false sense of security; not unlike what we currently have in regards to the MDA and the Iron Dome. The MDA cannot stop North Korean missiles and the Iron Dome cannot stop one or two nukes from eliminating Israel.

5. It’s actually unlikely that we really will solve the associated technical problems under any circumstances in our lifetimes. Of those problems facing it, the proliferation of space junk is the biggest one of all - and one that the program would actually contribute to rather than resolve. Every new satellite in orbit is one more thing orbiting the planet that needs to be tracked and that one satellite could transform into thousands of smaller pieces if it breaks apart (like a SpaceX satellite did just last month). Tracking 100’s of thousands of objects now is bad enough, but consider what happens when things start to blowing up in orbit (as they inevitably would given that space-based weapons platforms would become targets themselves). The systems that need to track all of that junk would become confused and overwhelmed quickly, immediately degrading the shield’s capability far below the needed 99.9%.

Cinematic visualization of The Kessler Syndrome, courtesy of the film Gravity (2013).

Pursuing space-based technology does carry some dividends just as it has always done with NASA, (the Heritage Foundation claims SDI gave us a trillion $’s in benefits, yet their calculations are dubious to say the least), but the ratio of benefits to spending is not nearly as effective with missile defense and as it is with NASA and if we’re not getting enough spinoffs and the shield will never work - what we are actually doing is throwing a lot of money into a black hole, never to be seen again. And just where is all of this money going to come from? According to the Trump Administration, who just asked for another $445 billion in DoD funding for FY 2027 (with a huge increase for the Golden Dome no doubt included), that money will come from cuts to Medicare and other social programs. Money for nothing doesn’t seem like a great bargain, especially when it costs us this much…

Copyright 2026, Digital Perspectives

A new definition of Insanity for the AI Age: Proposing the least realistic, most expensive solution possible to a problem that doesn’t actually exist and getting rewarded for it.

There are times when I feel that I may be the last sane human on planet Earth, the past few weeks has been just such a time. While it’s bad enough that we’ve had to listen to and take Donald Trump seriously now for 10 years, this particular rant isn’t about politics - at least not directly. What I’m referring to today is how AI Hype gone wild is somehow making the ridiculous appear credible and how this unreality is undermining not only the economy, but perhaps society itself.

It all started about a year or so ago when a handful of people started complaining about the rapid growth of data center construction due to AI Hyper-scaling; the concerns ranged from the growing costs of building them to high water usage and potential impact to power grids. While all of those are indeed legitimate concerns, the supposed fix to all of these problems mentioned wasn’t at all what one would expect - in fact - it seemed as though it were the least plausible solution possible; building data centers in space. At that time, not many were taking it too seriously, which of course is exactly what would one would expect when faced with such a hare-brained scheme. Then came Google, Amazon, Nvidia and of course Elon…

Elon Musk has stated that there is no more room on Earth for the rapid expansion of AI data centers and has vowed to put them into space. He argues that space-based, solar-powered AI data centers are the only way to scale computing power without overloading utility grids and exhausting resources on Earth. (feb. 2026)

Let’s Get this Straight

Before we talk about the science or economics behind the proposed solution, let’s stop for a minute to examine Elon’s (and the other tech giants’) rationale for proposing to put data centers in Orbit (or even the Moon as Musk has now done). What they’re saying is:

1. That making any progress with AI (now and forever) requires massive Hyper-scaling.

2. We’re running out of room on Earth to build giant new data centers (some the size of small cities).

3. We’re running out of power to operate and water to cool these behemoths.

4. That building these data centers has become too expensive and is definitely becoming unpopular.

5. So, to avoid having to fight for resources (land, water, electricity) with the stupid people of Earth who just don’t appreciate the need for Hyper-scaling, we’ll build our 5gw centers (which would be 30 million sq ft on Earth) in Orbit or on the Moon instead.

Elon Musk on Data Centers in Orbit or on the Moon: “The only logical solution…is to transport these resource-intensive efforts to a location with vast power and space. I mean space is called ‘space’ for a reason,’”

In response, CNN heralded the idea by saying “It isn’t as Crazy as it Sounds,” a title which should immediately alert readers - yet this was one of countless articles that did take it seriously.

Of Course, it is (Bat Shit) Crazy

Why? It all sounds so reasonable and realistic and those are the biggest tech companies and richest men in the world, they surely aren’t idiots, right? Well, besides the fact that their idea seems like a bad rehash of some B movie sci-fi plot (except that instead of the planet no longer supporting humans, it can’t handle data centers anymore), the economics of the idea make no sense whatsoever and the whole thing is also based on a huge misconception of what will be necessary to advance AI beyond its current evolutionary dead end of LLMs.

From Google’s Suncatcher webpage - “The Sun is the ultimate energy source in our solar system, emitting more power than 100 trillion times humanity’s total electricity production. In the right orbit, a solar panel can be up to 8 times more productive than on earth, and produce power nearly continuously, reducing the need for batteries. In the future, space may be the best place to scale AI compute. Working backwards from there, our new research moonshot, Project Suncatcher, envisions compact constellations of solar-powered satellites, carrying Google TPUs and connected by free-space optical links. This approach would have tremendous potential for scale, and also minimizes impact on terrestrial resources.”

For Reference - 5gw is enough power for either 500,000 homes. “5GW data centers are massive, proposed, next-generation AI infrastructure projects estimated to span 30 million square feet and cover areas nearly the size of Manhattan, such as Meta's proposed Hyperion site. These $100 billion, 5GW facilities are designed to house 2 million+ GPUs, demanding more energy than all current data centers in Northern Virginia combined.

So, essentially these morons at Google are saying, the Sun is free energy and there’s lot’s of room in space - I’m just paraphrasing. Even though they’re talking about putting data centers the size of Mount Everest in orbit and powering them with 1000’s of miles of solar panels - somehow that would be cheaper than building the same sized data centers on Earth where the vast majority of the planet is empty and where the entire surface is bathed in that self-same ‘free’ sunlight.

Of Course Nvidia already proved the whole thing by sending one GPU into orbit, didn’t they? Starcloud, an NVIDIA-backed startup, successfully launched the Starcloud-1 satellite in November 2025, featuring an NVIDIA H100 GPU to test AI processing in orbit. This initiative aims to establish massive, solar-powered orbital data centers, utilizing free space cooling and continuous solar energy to overcome terrestrial AI energy constraints, aiming for a 5-gigawatt facility. So one GPU proves the viability of a 30 million sq ft facility in orbit, hmmm.

For Reference - The International Space Station (ISS) consumes approximately 75–90 kilowatts (kW) of power to operate its systems, which is roughly equivalent to the energy needed for 40–90 American homes. Power is generated by 8 large solar arrays producing up to 120+ kW, with batteries storing energy during the 35-minute eclipse period of its 90-minute orbit. The ISS cost $150 billion to place into orbit and assemble and another $3 billion per year to operate and is now scheduled to be de-orbited because - guess what - it’s too expensive. China has a Space Station 1/3 the size of the ISS and Russia de-orbited their tiny station decades ago. No private companies have or are planning on putting any additional stations in orbit; (yet Google is planning for a 30 million sq ft facility up there - that would perhaps be about the same size as Babylon 5).

The ISS, which is as big as football field, in fact only has a Habitable Volume 13,696 cubic feet. There are 12 permanent EXPRESS Racks (EXpedite the PRocessing of Experiments to the Space Station) in the US Orbital Segment, which act as the primary, multipurpose payload shelving units, some of which can be used for computing and data processing. So $200 billion got us 12 racks with a few servers in the them.

How Much Would it Actually Cost?

Let’s run the numbers and I’m going to be generous and assume Low Earth Orbit (LEO) instead of the higher (and more expensive Geosynchronous orbit or the much more expensive Moon). To put a constellation of a million solar satellites into orbit to harvest all of that “free” solar power (both Bezos and Musk are proposing this) would cost about $2 trillion just by itself (assuming the average cost of around $1,500 per pound with maybe 30 satellites in each rocket adding up to a million satellites - each launch for the 30 costs up to $50 million). It’s worth noting that the Starlink satellites proposed for doing this are not designed at all to capture any significant amount of solar power. Here’s how they describe one of their “Constellations:”

Total Constellation Impact: The cumulative solar array surface area launched by SpaceX for Starlink is massive (9,422 satellites), estimated to be over 50 times the total solar panel area of the International Space Station. (While exact, single-satellite generation figures are proprietary, analysts estimate that with thousands in orbit, the total constellation capability is immense). Translation - it takes nearly 10,000 of Starlink’s satellites to generate 50x the tiny amount of solar power generated by the Space Station.

This is why Musk said he’d need a million satellites; but even that figure is likely low and of course none of that includes the additional cost of building a city-sized data center in orbit. It’s hard to find any rubrics here, but I located one construction site that mentions the need to have 55 pounds per sq ft, so let’s say we multiply that by just a 1 million sq ft (instead of the 30 Google was hyping) and we get 55,000,000 pounds and if we multiple that by $3,000 per pound (to include assembly cost) we get $165 billion for 1 data center. Now this is likely a massive underestimate and doesn’t get us close to the 5gw scale facility Google is dreaming about either (nor does this address whether most of the technological challenges with actually doing it have been figured out, because they haven’t been - so it might not even work).

For Reference - Building a 1-million-square-foot, multi-megawatt data center typically costs between $600 million and $1.1 billion+, based on industry averages of $600–$1,100 per square foot. Costs are driven by high-density power requirements (40-50% of budget), specialized cooling systems (15-20%), and robust, secure construction.

So, a typically large data center would cost more than 100x more to build in orbit than on Earth and the solar power array would cost perhaps a 1,000x more - making it the most expensive “free energy” in history. This is just back of the napkin quick calculations, but obviously there are countless folks involved in the aerospace industry who are quite familiar with the costs and difficulties of building in space. Like CNN though, most of the major news outlets seem to keep saying - hey this isn’t actually crazy - it’s an “audacious solution” to all of our current AI-related woes, (in fact, the “audacious” quote has shown up so often now that it must be straight off a PR release).

Why Does This Matter?

Well, it’s bad enough that AI Hype has now swallowed the majority of the US economy and is leading to 100’s of thousands of real people losing their jobs every few months (either through anticipated AI replacement or to fund the ever-growing AI CAPEX Black Hole), but now these same companies who want you to accept without question that they (and you) actually need all of these data centers - also are asking you to completely turn off your brain and let them claim that any sci-fi fantasy they imagine is already reality and demand that you believe that too. It’s a lot, in fact it’s way too much - it’s the death of reason itself. And now we get back to the politics part of the rant - it’s not that much different than Big Brother telling you that 2+2=5 and having you not only believe it but love him for it too - and Big Brother is alive and well and living at the White House and Mar a Lago.

Data Centers in Space is about a million rockets too far; it’s not our problem that Bezos and Musk can’t figure out how to make money with either their AI or their rocket companies - but the absolutely last thing we need to have happen is get stuck bailing them out for the next 100 years after they go broke throwing money into a Black Hole. They will take the economy down with them and perhaps we’ll all learn then that stupidity and business do not mix well.

Copyright 2026, Digital Perspectives

1. Because AI is a lousy writer. Yes, it can format pretty documents and even do citations (which is where I might just crack one day), but even when copying the style of actual writers, the output leaves me less than impressed. This may get a little better over time, but I doubt if it will ever progress from adequate writing to great writing. And after spending decades trying not to be mediocre, it would just be too depressing to give in now.

2. Because somebody real has to help train these things. Leaving the next generation of AI models to the mercy of Grokipedia or Reddit or Pravda’s disinformation factory is just a recipe for ending up with a real-world Terminator (if you don’t count Arnold as Governor as already having achieved that). My content is the only thing that might keep these unhinged models from going overboard. I’m saving the world, one post at a time.

3. Because I’m getting old and my brain might shrink. Yes, this is a real thing (and it can happen even without AI - for example by watching TikTok or too much sports). I hate Sudoku, so writing it is. Also, I’m becoming fairly convinced that the LLMs are crazier than I am.

4. Because someone must be around to teach the next generation of humans (after the GPT generation self-destructs). As we pull ourselves out of the Trump Dark Ages, where millions relinquished their ability and desire to think, the few - the proud and the very old writers that remain will spark the next enlightenment.

5. Because I’m making it my personal goal in life to prevent anyone from becoming a trillionaire (just on principle). The greatest irony of the AI Age will undoubtedly be that some of the dumbest people in the world (yes, we’re talking about you Elon and Sam) became some of its richest ever. The last thing I want to do is to facilitate their path to riches by legitimizing their (currently) ridiculous products. I’m afraid I may be too late to stop it though.

Here’s some tag lines / potential ads for today’s most popular Use Cases for LLM-based AI:

• Cheat your way to higher grades and career success!

• Got Plagiarism, hell yeah!

• Thinking is so retro…

In all seriousness, I don’t intend to use LLMs (or any AIs) for any of the things that a) I’m good at or b) that I really enjoy doing. This situation is a little bit like having a robot play tennis for me - why would I outsource my fun? I don’t want LLMs to write for me because I know the writing quality would suffer and that I’d lose the very skills that I was trying to exploit or conserve (and then it’d be someone else’s creation, although maybe a tiny amount would still actually be mine - no royalties of course).

Weird Al’s Amish Paradise.

Copyright 2026, Digital Perspectives

Thirty years later, that hoped-for Democratization is turning into a frightening, all encompassing Techno-Feudalism that not only won’t deliver on new forms of empowerment, but seems likely to strip away all of the existing Democracy and personal empowerment that we had before. Immense wealth and nearly unimaginable inequality are emerging from Silicon Valley as 100’s of thousands of actual tech workers get laid off every few months. AI tools like Grok are allowing users to strip the clothes of women and children, while Social Media has helped to topple Democracies across the globe and facilitate human trafficking and worse. Products and Services that were supposed to empower us are now being used to help exploit and terrorize us.

Techno-Feudalism is a contemporary socioeconomic theory proposing that capitalism has evolved into a new, neo-feudal system where dominant tech giants (the "lords") control digital platforms, data, and algorithmic infrastructure. Instead of profit-driven market competition, these platforms extract "digital rent" from users and businesses ("serfs") who rely on them to function. (Google’s AI Overview definition, which seems somehow appropriate)

The Insanity of 2026

This year marks the official turning point whereby we’ve shifted from the edge of the cliff to the swan dive into the abyss. This week for example, Elon Musk just became $90 billion richer even after announcing that he’s giving up on the one business of his that was actually making decent revenue (Tesla) and that he was merging two other companies that weren’t earning much or anything. Let’s take a moment with Musk because while there is a lot of craziness happening in big tech; he has been at the center of much of it for quite some time and seems to have profited the most from it:

• In 2025, after making himself nearly universally unpopular (first through how he ran the unprofitable Twitter / X, then through the 2024 election and then DOGE), Musk managed to kill the Tesla car brand.

• Having recognized that, over the past week he announced that he was going to stop making cars and switch to robots that no one has seen let alone bought yet, because of course Tesla wasn’t really a car company (like X wasn’t really a social media company), they were both AI companies - or now AI / Robot companies. And of course with each bullshit announcement, the fanboy investors rush to pump more money into his never-ending Ponzi scheme empire.

• He also announced that besides Robots, that Tesla would be churning out Robo-Taxis - despite the fact that none have really run successfully anywhere and that his self-driving software has never worked and he refused to install the sensors and cameras on the cars necessary for them to ever work (to save money).

• It’s worth remembering here that Musk was not involved in the initial designs for the Tesla models that did achieve success and the one new model that he did produce (after 15 years of sticking with the old ones and 5 years or more of promising a pickup) was the CyberTruck - this turned out to be the worst lemon since the Edsel and he can’t give them away now (he sold 100’s or thousands of them to SpaceX in another circular deal).

• So, the extra $90 billion to his personal wealth came from the announcement that xAI (which has never made more than $500 million) and SpaceX (which made $15.5 billion in 2025) would merge before the SpaceX IPO. So, this new combined company that made $16 billion last year, has reached a valuation of $1.25 trillion (which is 78x the revenue). Some of the optimism here is related to the more or less insane $150 billion Golden Dome project (which we will write about soon) which many expect to heavily favor Musk’s companies when it comes to awarding contracts, but even still - it’s a sky-high valuation that no defense contractor has ever seen before.

This is actually just a small taste of how Musk and Big Tech at large have morphed into a perpetual money making machine that seems mostly now to be divorced from the economic reality that everyone else inhabits.

The God Kings of Tech

We’ve got a dozen or more mega-wealthy tech barons (everyone from Sam Altman to Larry Ellison) who also now to seem to have determined that their Pharaonic-level wealth has indeed made them into the God-Kings of society. And as God-Kings, many of them have decided to try to remake society to better suit their ambition to remain God-Kings and of course Pharaohs have never tended to hold a particular affinity for Democratic institutions anyway. Much of the money and support for America’s current authoritarian transformation is being directly underwritten by these same individuals; whether it be through contributions to Trump’s Golden Ballroom or Triumphal Arch (the new, gilded White House is becoming the symbol for this Digital Gilded Age), or a $28 million bribe to Melania for her vapid documentary (the highest salary ever paid to someone as part of documentary and a near record Hollywood talent salary in any context).

During the first Gilded Age, the Robber Barons of industry flaunted their wealth and encouraged or sponsored corruption, but never at this scale or with this level of disdain for the rest of us. And in many ways, those tycoons understood or appreciated their workers in a way that our current Gilded Age moguls never will. The goal of the original Robber Barons was to keep the workers in line, not to eliminate them - which sometimes led to finding ways to better support them (although not always). The current generation of Gilded Tycoons simply view workers as irrelevant - something to be replaced - which they are trying their best to accomplish as soon as possible. From their perspective, any of the economic activity associated with actual workers is already immaterial - in other words - the “leg” of the K-shaped Economy could be amputated without impacting their ever-soaring valuations & wealth.

Tony Scott’s Enemy of the State in 1998 portrayed some of the dangers associated with emerging technology (at least in regards to the surveillance / privacy). The movie viewed the danger coming from Government, which has turned out to be only a small part of our new reality in 2026.

Absolute Wealth is Power…

Power the likes of which our previous Gilded Age Tycoons could have only dreamed of; not only because of the volume of cash now involved, but also due to the incredible new ways that power can be exercised given the massive of authority that’s been shifted to the Executive Branch coupled with the technologies owned by the Digital Robber Barons’ companies. We are witnessing first-hand, almost on a daily basis, that the hyper-concentration of wealth is leading to an unprecedented lurch backwards for society - one that could actually send us right back into a Digital version of the Middle Ages, with power and economic security being stripped away from nearly all of us. We are in danger of becoming the new Serfs; this isn’t like the Industrial Age where workers could push back against the Robber Barons through strikes - today’s industry is on the verge of not needing any workers. This will certainly happen if most of us continue to be more or less silent about it - we still have some power left, but if we fail to exercise it will simply fade away (along with much else).

Copyright 2026, Digital Perspectives

I have always been fascinated by the future - but it wasn’t just me; I come from a generation where the promise of an ever-greater future was a key part of the bargain for everyone living in America. I grew up watching man go to the stars - I myself got to see the last Apollo take off (in 1975) and much later I was able to take our son to see the last Space Shuttle launch (in 2011). The future for me has always held the promise of better, brighter days to come - of continuous progress towards new discoveries and capabilities - all of which in turn would make our lives that much more interesting.

Visions of Progress…

The view that I held was the one predominately portrayed in the media and in our popular culture (except in film, but disaster films and dystopian futures have always been quite popular as entertainment). I came across some very interesting clips on YouTube from the 1960’s and even earlier as well from Walt Disney himself. If I had to characterize this vision of progress further; I’d say that it tended to include the following attributes:

• A focus on Science (and how Science and Industry would work hand in hand to solve the world’s challenges).

• An expectation that the human condition would continue to improve (in an unbroken chain of progress).

• That the world was coming closer together; as a result both of communication technologies and global organizations dedicated to bringing nations together in common cause.

• A focus on the wonders of technology and how these new-fangled devices would change our lives.

Walter Cronkite (late 1960’s) visits the home of the future and discovers “E-learning,” among other things.

Walt Disney introduces an EPCOT that never was. If only he had lived long enough to build it, it may have indeed changed our future.

Futurama - The 1939 World’s Fair: Many World’s Fairs were built entirely around visions of the future; although this one seemed to be filled with superhighways (and it came true too).

The Future, Dystopia Now?

There has always been an alternative, cautionary narrative of the future; going back all of the way to silent films (and some novels older than that) through to today. Dystopian futures have always been entertaining; as the new season of Fallout has just clearly demonstrated, but they never seemed quite real - at least for the most part. There were the times in the 1960’s and the 1980’s where the threat of nuclear war loomed over us like a pall, but that passed (or did it)? And then came the 2020’s with the chaos of Populist governments (including perhaps the worst one here) and the devastating impacts of a Global Pandemic - all of sudden the notion of entering a Dystopia seemed more realistic.

Things to Come (1936) - HG Wells’ dystopian vision of the future, brought to the silver screen.

Futureworld - not the one from EPCOT (although it really seems as though they tried to re-imagine EPCOT in this film).

What The Experts Thought

Did anyone see - this - coming? Not really. Well, much of what they predicted did come true of course, at least within the narrowly-defined view of technological progress.

Walter Cronkite interviews famous futurists in 1967…

Arthur C. Clark predicts the future (in 1964) - in this video he predicts “Digital Nomads,” although commuting and cities still exist. Oh, and his prediction of the Planet of the Apes via cybernetic implants is pretty far out there (although Elon tried that and it didn’t pan out). No wonder this guy wrote; “2001, A Space Odyssey.”

A New Reality

I have begun to question whether the future will be better than the past - that progress will continue in a beneficial manner for the human race and the planet. Why do I say that? Well, here’s a short list:

1. Vast numbers of people and politicians have chosen to reject Science in favor of rabid belief in superstition and disinformation (which in reality are the same thing).

2. Reason and common sense have been too often ignored and set aside to be replaced by hate, ignorance and intolerance - and this transition has been accepted in dozens of societies across the globe.

3. Too many of the current waves of progress have disregarded the potential for harm. While many previous innovations have eventually proven dangerous or disruptive, it seemed as though these were balanced by others that were indeed beneficial (and there were efforts to correct the problems later). One gets the impression today though, that all technologies, whether harmful or not to humans or the planet will be promoted now not due to their potential benefits but rather solely for their profit potential, (to an ever-shrinking pool of mega-rich people or corporations).

Perhaps I’m getting cynical due to the almost daily chaos we’re suffering through now in the early days of 2026. But why is it that we find ourselves in the midst of such chaos? Had more people rejected the voices of Hate, Intolerance and Ignorance in 2024, we wouldn’t be facing the consequences of the world being led by the whims of an 80-year old tyrant. This isn’t how the country, the world or the future is supposed to work…

Copyright 2026, Digital Perspectives

Will widespread replacement of human workers by AI capabilities lead to an economic collapse?

Why Would That Be The Case?

Let’s expand the question a bit - if far fewer humans are required to work, how will they live and furthermore - how can they support economic development or growth. In particular, we’re referring to consumer spending - according to most sources, this spending drives somewhere in the neighborhood of %70 of GDP. Out of that, the vast majority of spending is done by people considered to be in either the Middle Class or Upper Class. Middle Class and Upper Class jobs, (read - “White Collar”), are precisely the ones now most at risk of AI replacement. There is some dispute as to how much AI replacement is actually occurring because there are no associated standards for reporting on such actions (by employers); we’re getting the news on this more or less second-hand. For example, many pundits are extrapolating that AI replacement is hitting college graduates particularly hard to due to the rapid and steep decline in openings for them in 2025 (and that’s accompanied by at least some anecdotal evidence - but no official statistics in relation to AI that is).

It’s worth noting here that automation has historically replaced large groups of skilled workers over the past several centuries, (across a wide variety of industries); so in one sense, this is nothing new. What might be new this time is that all of the various automation trends have likely combined now to eliminate most work opportunities going forward. With previous automation cycles, other types of work always emerged to provide alternative employment opportunities. For example, as factory work became more automated, (eventually by using robots and now robots equipped with AI), many people moved into the IT or Service Industries. The IT path offered comparable wages and benefits, the Services Industry typically did not. But now with the recent advances in AI, jobs in both the IT and Services Sectors are being replaced on what’s likely a massive scale (everything from programmers to drive-thru window attendants). And so far, no AI-generated jobs are popping up to replace what’s being lost - in fact most “AI” roles in IT are just folks who have rebranded themselves as their old roles went extinct.

In other words, where are the all new jobs that AI is supposed to create or will otherwise emerge to balance the job losses? The hundreds of billions of dollars (probably trillions) in data center investments result only in a brief surge in construction, but very few ongoing jobs. So, the equivalent of a trillion $ + stimulus may end up only creating a few thousand jobs in the US (instead of millions), while subtracting millions of existing jobs.

What About Universal Basic Income (UBI)?

Well, first of all, there isn’t any and secondly - welfare programs that are similar to this have nearly gone extinct in the United States, (in fact, Welfare is considered a dirty word here and has been for a very long time). Besides being dead on arrival politically, where exactly would the United States get the money to pay for UBI? (considering that we’re now $38 trillion in debt).

The yearly cost of a U.S. Universal Basic Income (UBI) in 2024 estimates ranges significantly but typically falls between $3 to $4 trillion annually for providing $1,000 per month to all Americans, with some proposals hitting $4.7 trillion or more, requiring substantial tax increases or major cuts to existing programs.

Oh, and who exactly in the US is actually able to live on $1,000 a month anyway? The answer is no one and that means that even with say $4 trillion in UBI spending each year, that the vast majority of Americans without jobs and just UBI wouldn’t have any disposable income meaning that the nation’s GDP would come crashing down (which in turn makes those UBI payments less affordable as tax revenues plummet). And, no the AI companies won’t be able to pay even a part of a UBI because none of them (with the exception of Nvidia) are actually making a profit and that doesn’t seem likely to change (and of course we’re giving them all massive tax credits and abatements).

Where Will Lead Us?

And speaking of profits, one of the great ironies emerging in this story is the notion that the primary ROI path that seems to be emerging for AI companies is precisely the one we’re referring to here - replacement of as many human workers as possible (because few other revenue streams are emerging). Although, much of the ROI in that context may not even translate back to the AI companies themselves unless they create so much phony wealth (through Roundtripping) that they can gobble up many other traditional companies. The path that we’re on now, (which is barreling along like a freight train without much forethought), seems to be leading us to the following outcomes (within the next 2 to 5 years):

1. Massive increases in structural / long-term unemployment in the US.

2. Significant drops in GDP beginning in a about a year or two as disposable income begins to dry up.

3. A significant reduction on the size of the Middle Class and increases in the amount of poverty and homelessness.

4. Additional socio-political unrest as income inequality gets far worse than it already is.

5. Potential budget collapse and national bankruptcy.

Yes, this may sound like “Doomer-style” predictions, but what other outcomes could we expect if as many as 20 million jobs are lost (this is only around 10% of all US jobs), no UBI is instituted and GDP drops around 3-5%? It may actually be a conservative prediction of the disruptions that lay ahead if we remain on our present course. No one is taking this seriously - yet. By the time we do take it seriously, will there be enough runway to course correct? Who knows? But one thing that is clear, the continual removal of human workers from the economy will eventually shrink that economy - there’s no getting around it. Humans spend money, AI (agents, etc.) don’t.

Copyright 2026, Digital Perspectives

The Provisions of Pillar 2

1. Create Streamlined Permitting for Data Centers, Semiconductor Manufacturing Facilities, and Energy Infrastructure while Guaranteeing Security. This was covered by the AI Infrastructure Executive Order (EO).

2. Develop a Grid to Match the Pace of AI Innovation. This provision goes a little beyond what’s in the AI Infrastructure EO. While it’s not a bad idea to improve the nation’s power grid, the question is one of priorities. The current grid is having problems supporting existing power demands (particularly in Texas). The implication here is that the power grid in future will be optimized and / or prioritized to support AI datacenter power needs (potentially at the expense of more traditional needs). Here’s a quote - “By stabilizing the grid of today, optimizing existing grid resources, and growing the grid for the future, the United States can rise to the challenge of winning the AI race while also delivering a reliable and affordable power grid for all Americans.” In this quote, Americans come second to ‘winning the AI race.’

3. Restore American Semiconductor Manufacturing. This one is a little perplexing given that we’ve already got a lead here and at the same we’re stating this, the CHIPS Act is being actively undermined. There’s nothing wrong of course with promoting the semi-conductor manufacturing here in the US, but it is rather odd to see language like this included in the Strategy: “The Trump Administration will lead that revitalization without making bad deals for the American taxpayer or saddling companies with sweeping ideological agendas.” The implication may be that even in areas where the current and prior administration agree, they have to try to tear down any of the prior achievements or positions - just because. Overtly politicizing national technology strategy is not generally a great idea. And BTW - they go on in the plan and Woke AI EO to introduce a massive sweeping ideological agenda

4. Build High-Security Data Centers for Military and Intelligence Community Usage. Well, of course we’re already doing this and have quite a few of them. We may need more, but then that begs the question about AI and Military Doctrine - and by the way we haven’t actually reconciled our Doctrine with AI yet. The first sentence in the provision here is a little troubling because of the ambiguity involved - “Because AI systems are particularly well-suited to processing raw intelligence data today, and because of the vastly expanded capabilities AI systems could have in the future, it is likely that AI will be used with some of the U.S. government’s most sensitive data.” Raw Intelligence in this paragraph could mean intel on American Citizens, not just foreign adversaries. We’ve already been headed down that road for 20 years, it doesn’t just feel great thinking that the journey may now be turbo-charged by AI.

5. Train a Skilled Workforce for AI Infrastructure. This provision relates to what I talked about in my article on What AI Skills? You see, there are no AI-specific skills or jobs connected to this - only Data Center skills and roles. Almost everything called out under this provision already exists - through partnerships with Community Colleges and school systems and even with things like Apprenticeships. The problem is that much of that was being managed and spearheaded by the Education Department - the same Education Department that’s being slashed to bits. OPM, the other big agency supporting these types of things is being slashed as well (by 70%). Good luck getting this one done.

6. Bolster Critical Infrastructure Cybersecurity. Yes, absolutely, positively - but wait; not really. DHS, the lead for this provision, is in the process of having its Cyber workforce slashed as well. So, I think they were just kidding on this one, too bad. Cyber Security is also being cut across other agencies as well - it’s just that DHS was charged with overseeing critical infrastructure, so cutting its Cyber capabilities precisely when they’re most needed isn’t likely to turn out well.

7. Promote Secure-By-Design AI Technologies and Applications. This sounds great on the face of it. By all means, let’s do this! Although, if we think about it a little - AI Assurance (as they call it) can really be thought of as a subset of Cyber Security and Cyber Security is - um, being cut. There’s a bigger problem with this one though - AI itself (whether assured or not) represents potential risks that we don’t fully understand. Yet, we’re on the cusp of handing off autonomous or at least semi autonomous tasks to ChatGPT-derived models. Giving FPV drones in Ukraine autonomous capability is one thing, plugging into our C2 (Command & Control) systems is another. This is the most important part of the plan that’s actually missing.

8. Promote Mature Federal Capacity for AI Incident Response. This one is a good idea and more or less attacks the real AI-related risk indirectly (e.g. ‘risk’ here being defined as critical systems somehow going down due to non-specific AI failures). But honestly, this provision is more or less adding some new steps to existing techno-bureaucratic processes. It kind of begs the question, though as to how AI system incident response actually differs from traditional incident response. Maybe, it’s because in case of an AI system, (we might call it HAL), the system gets angry and decides to cut your air hose or lock you out of the ship? Hmmm, how to do we capture that in the incident response protocol? Got it - we blow the hatch and disengage HAL’s higher functions - but oh, he’s on the Cloud now - crap.

This part of the AI Action Plan isn’t particularly well thought-out and seems skewed more towards Big Tech than the American people. BTW - it’s worth noting that I’m pretty sure that there was no request for industry or public input or review of this plan and there’s been very little discussion about since it came out either.

Copyright 2025, Digital Perspectives

1. What AI Skills?

2. The AI Executive Orders, part 1

3. The AI Executive Orders, part 2

4. The AI Executive Orders, part 3

But the National AI Action Plan goes beyond what’s in the 3 Executive Orders and includes more than the missed opportunities highlighted in the What AI Skills? article. Today I’m going to look at the remaining provisions of the Plan not already covered previously. I’ll review them by the “Pillars” (sections) listed in the report:

The new AI & Autonomous Systems Virtual Proving Ground?

Pillar 1 - Accelerate AI Innovation

Hidden within the plan are few positive suggestions, doubtless leftovers from longer-term discussions. Some of those from Pillar 1 include:

• Encourage Open-Source and Open-Weight AI - This is focused primarily on research efforts and Academia (and whatever smaller companies haven’t already been gobbled up) and helps to level the playing field.

• Invest in AI-Enabled Science - One of, if not the only, areas where the current Administration has pledged to invest in rather than cut Science. Unfortunately, you can’t cherry-pick a few Scientific nuggets to focus on while decimating the rest w/o impacting the area that you happen to like.

• Accelerate AI Adoption in Government - This would be good if weren’t already happening in such a way that’s actually weakening Government. I talked about this earlier this year at length in regards to DOGE. BTW - ‘Data Science’ is not an AI Skill, it’s Data Science (same goes for Software Engineers). This goes back to the discussion about how there aren’t any real ‘AI skills’ and these others being mentioned are actually being endangered by AI (Software Engineering may actually be the single most endangered one).

• Advance the Science of AI - This may be the one place where they acknowledge that LLMs aren’t the end all be all of AI.

• Invest in AI Interpretability, Control, and Robustness Breakthroughs - Again, this is likely a leftover from earlier times - it makes sense.

Now the Problems in Pillar 1

• Combat Synthetic Media in the Legal System - On the face of it, this one sounds really good, but it’s missing something; actually 2 somethings - a) it doesn’t address Fair Use Copyright theft by AI companies and b) it doesn’t address the impacts of use of Synthetic Data. In fact, elsewhere both of these concepts are being promoted by the Administration (e.g. using Fair Use to steal Intellectual Property as well as using Synthetic data for training). Both of these issues are worse than the one that is being addressed. Why? Because, companies that are worth Trillions of dollars are stealing the work of ordinary Americans and getting a pass on those thefts (we’re going to have an article dedicated to Fair Use soon). As far as Synthetic Data goes, the problem is more subtle. Long-term use of and dependence upon ‘fake’ data to evolve new generations of models is likely going to make them even less predictable and understood. Combined with the current focus on the ridiculous scaling up of data centers it’s a good bet that this will wipe any advantage our models will have (globally) in even the near-term.

• Ensure that Frontier AI Protects Free Speech and American Values is embedded in the Woke AI EO, Remove Red Tape and Onerous Regulation is covered in “Streamline AI” EO, Empower American Workers in the Age of AI (I covered this in my What AI Skills article),

• Build World-Class Scientific Datasets - This is related to the Synthetic Data problem, but can include real data. What’s missing here is; a) a better context to determine what they mean by this and b) more importantly deflects away from new model design - again assuming that the end all be of AI development will always be LLMs. Yes, good data will be required for any AI models, but it’s important to make the connection that it’s not all about feeding the model; the new generation of models (the ones that will take us from where we are to AGI) will need to be architected, not just “fed.”

• Support Next-Generation Manufacturing - This one worries me. I think it’s code for ‘let’s replace more people with robots;” (Led by DOC through NTIA, convene industry and government stakeholders to identify supply chain challenges to American robotics and drone manufacturing), but it’s sufficiently vague on that and also does call out Chip manufacturing.

• Build an AI Evaluations Ecosystem - Ok, this one reads pretty well on the face of it - except, (and this is a rather big exception), it’s not making any reference to the context for the measurements. The EU had a similar problem with their recent guidance and they defaulted to some relatively meaningless measures like computing power. What should have been called out here are a couple of things; a) the link to basic science to figure out what it is we’re supposedly measuring, and b) the development of any guiding principles to peg those measurements to.

• Protect Commercial and Government AI Innovations - This should have been one of the most important (and thus most detailed) provisions. It refers to the intersection of AI and Cyber Security and is almost completely missed here. I’m not sure how we’re going to safeguard any of our AI if we’ve giving it this little consideration (while at the same time cutting our Cyber budgets - that’s a question for another article).

• Drive Adoption of AI within the Department of Defense - This begs the question, why are we calling this out separately from Driving adoption in Government? As a rule, we shouldn’t view the DoD as if it’s a government unto itself (for a lot of reasons). The provision for “AI & Autonomous Systems Virtual Proving Ground” is made without having adding any discussion on risks for Autonomous systems. Not good.

In part 2 of this series, we’ll look at Pillar 2 in the National AI Action Plan.

Copyright 2025, Digital Perspectives

Problem #1 - (Lack of) Vision

I’m not sure whether this is worse on some type of rating scale than the other issues, in fact, I think it might be comparing Apples and Oranges. But this problem stands out to me as the most immediate because it’s the trigger for all of the other problems. This EO envisions that growth and evolution of AI in the US must necessarily follow the current path of “scaling up and out” ever-larger data centers to ‘grow the models.’ What’s wrong with that you ask? Here’s a just a few reasons:

1. It seems to forfeit or dismiss other, better evolutionary paths (none of which are even mentioned in this or the other EOs and the National AI Action Plan).

2. This implies that we’re placing all (or most) of our bets on a technology that we don’t actually understand and that’s likely to be left in the dust by other, better approaches to AI design and architecture. We’ve talked about some of that here before…

3. The current generation of LLMs will never achieve AGI - this EO seems to leave us stuck on and with LLMs as the end-all, be-all of AI development and adoption. This is beyond short-sided.

Add all of this up and we’ve got a pretty good formula for losing the AI Race; and losing it while encountering the all of the harmful impacts of the associated problems. To be fair, this lack of vision didn’t start with the current Administration, but they have totally bought into it. The Groupthink associated with this stunning lack of imagination is going to be the number one reason why the US doesn’t win the AI race as it is will continue to redirect money and efforts into ever bigger data centers as the AI market further consolidates, locking out all of the innovators.

Problem #2 - Environment

Facilitating the rapid build-out of mega AI data centers (the AI only references data centers costing more than $500 million) is a recipe for disaster. The EO also calls for building these on public lands (and curiously it also mentions superfund sites as well). Most public lands in the US are out West, where we are already suffering enormous water shortages that are only predicted to get worse. Why are we choosing to accelerate the problem - this kind of action combined with other massive water wasting (like Arab countries using the last aquifers out West to grow Alfalfa) could turn much of the American West into Ghost Towns in 25 years or less. Not convinced, check out this study on global fresh water loss. Much of this EO is crafted to ensure that any & all environmental reviews are bypassed, especially those that would look at water issues.

Problem #3 - Energy

The EO also calls out data centers that are using at least 100 Megawatts MW. For comparison, 100 MW could power up to 50,000 homes. Many of the data centers being planned will consume much more power than this. In order to meet this growing energy demand, coal plants that were scheduled to close are being kept open (adding more carbon to the atmosphere) and new nuclear plants are in the process of being built (when we still haven’t figured out what to do with waste from the old ones).

BTW - they went out of there way here - not to mention - renewable energy sources such as Wind and Solar (although they might qualify as other dispatchable baseload energy sources, whatever that means):

(ii) natural gas turbines, coal power equipment, nuclear power equipment, geothermal power equipment, and any other dispatchable baseload energy sources, including electrical infrastructure (including backup power supply) constructed or otherwise used principally to serve a Data Center Project.

Other Problems

There are a few other issues associated with this EO that need to be highlighted; they include:

• This isn’t just about streamlining approval, the EO actually includes a provision for Federal subsidies of companies (some of the wealthiest in the world) to pay for data center development. Sec.  3. Encouraging Qualifying Projects.

• Revocation of a similar AI EO from the Biden Administration because it emphasized “Clean Power.” Sec. 4. Revocation of Executive Order 14141.

• A careful review of all of the ‘streamlining’ provisions in the EO shows a comprehensive elimination of any and all oversight for the approval (and later management) of such data centers including tossing out provisions about toxic waste and endangered species.

What’s perhaps the worst thing here with this EO is that all of the negative outcomes associated with continuing down this path are completely avoidable. In other words, not only could we win the AI race, by not doing this - that is in fact the only way we’ll win it. Instead, we’re going to lose the ‘race’ while getting stuck with an environmental nightmare and a failing national energy grid.

Copyright 2025, Digital Perspectives

• Led by the Department of Labor (DOL), the Department of Education (ED), NSF, and DOC, prioritize AI skill development as a core objective of relevant education and workforce funding streams…

• Led by the Department of the Treasury, issue guidance clarifying that many AI literacy and AI skill development programs may qualify as eligible educational assistance under Section 132 of the Internal Revenue Code, given AI’s widespread impact reshaping the tasks and skills required across industries and occupations…

• Led by the Bureau of Labor Statistics (BLS) and DOC through the Census Bureau and the Bureau of Economic Analysis (BEA), study AI’s impact on the labor market by using data they already collect on these topics, such as the firm-level AI adoption trends the Census Bureau tracks in its Business Trends and Outlook Survey. These agencies could then provide analysis of AI adoption, job creation, displacement, and wage effects.

• Establish the AI Workforce Research Hub under DOL to lead a sustained Federal effort to evaluate the impact of AI on the labor market and the experience of the American 7 Executive Order 14277 of April 23, 2025: “Advancing Artificial Intelligence Education for American Youth,” Federal Register 90 (80) 17519, www.govinfo.gov/content/pkg/FR-2025-04-28/pdf/2025-07368.pdf. 8 Executive Order 14278 of April 23, 2025: “Preparing Americans for High-Paying Skilled Trade Jobs of the Future,” Federal Register 90 (80) 17525, www.govinfo.gov/content/pkg/FR-2025-04-28/pdf/2025-07369.pdf. 9 Revenue Act of 1978, 26 U.S.C. § 132. 6 AMERICA’S AI ACTION PLAN worker, in collaboration with BLS and DOC through the Census Bureau and BEA…

• Led by DOL, leverage available discretionary funding, where appropriate, to fund rapid retraining for individuals impacted by AI-related job displacement…

• At DOL and DOC, rapidly pilot new approaches to workforce challenges created by AI, which may include areas such as rapid retraining needs driven by worker displacement and shifting skill requirements for entry-level roles…

Basically then - “Workforce Empowerment” in this plan consists of; a) retraining in AI, b) studying the impacts of AI on the workforce and maybe recommending to do something about it and c) probably engaging in AI retraining when it’s determined that AI is taking all of the White Collar jobs. On the other had, if we were to view the top tech CEOs of AI-related companies as the workforce, the rest of the plan seems geared towards ensuring that one or more them becomes the world’s first trillionaire (we’ll talk about that more in the follow-up article).

What’s wrong with this view of Empowerment?

Isn’t this the way we’ve always viewed new technology trends? Yes, it is. Unfortunately, Artificial Intelligence is unlike all of those previous trends for one simple reason - there’s no new skills required as a result of it. If anything, AI adoption represents the biggest ‘deskilling’ in the history of labor as more and more supposedly human activities are made redundant and costly artifacts of the distant past. In effect, there is nothing in this Action Plan that empowers workers because there is nothing that they can be taught to do that they don’t already know.

What about AI Engineering & other new jobs?

Well, aren’t there going to be a slew of new AI-related technology jobs like AI / Model Engineers given the enormous investments being made? Well, no - there won’t be. Why? Because, a) the industry is consolidating among an ever-shrinking cadre of mega-companies that will likely control 90% of the funding and most outcomes associated with AI and b) AI tools will be used more frequently to design AI until the point where human engineers become redundant - this is going to happen much faster than most realize. There have been some associated AI jobs related to the training process for models where folks help to ‘fine-tune’ results - but those which have been largely outsourced anyway, will also become unnecessary sometime soon. Are there any other new job roles being created? Nope. I’ve seen articles where folks have tried to make them up - this process usually involves adding the letters “AI” before a dozen or more existing job titles. Again, the process will be working in reverse with titles disappearing rather than being added.

Is this Empowerment in a personal sense?

Well, yes - many of us will be able to do more work than before and also get involved with more types of work that we’ve never done before. The question will be whether anyone wishes to pay us for that work. This new economy might necessarily become informal and more focused on personal businesses with fewer and fewer employees required in medium to large-sized companies. And importantly, in the Government’s AI Action Plan, there is no mention of this, nor any attempt to help out (other than perhaps in conducting a handful of studies that will likely be ignored by this administration). There is zero funding dedicated to compensating for AI-related job losses or paying for the non-existent AI skills training.

So, what are AI Skills anyway?

Anyone who has used ChatGPT, Copilot or any of the similar tools now available already knows the answer to this question: an AI skill is simply the ability to provide input to one of these tools and ask it to perform the skill on your behalf. So, if you can communicate in any way (through writing or speech), you’ve already got the skills necessary to use most of these tools. More specific AI tools, such as those used to auto-generate code, require some understanding of the coding process - but not much and these of course are existing skills, not new ones. Yet, despite the obvious fact that there are no AI Skills to be taught, both industry and government leaders continue to insist (quite counter-intuitively) that there are such skills. And these skills are not only referenced (yet never specified) again and again, they’re held up as the only answer to the wave of AI-induced job losses. Not good.

We’re stuck in a “Don’t Look Up” doom loop of ignoring the impending Great Replacement while being told that if we only learn to tilt our heads in the other direction we’ll be saved. What would Workforce Empowerment in the Age of AI actually look like? That’s a tough question - but it’s one that our leaders are skillfully ignoring in a race to adopt AI with zero restrictions and forethought.

Copyright 2025, Digital Perspectives

I had expected something like this to happen in my lifetime, but not this quickly and not in the manner that it has actually occurred.

Let me step back a moment and provide some of my perspective on the matter. In the late 1980’s I first became involved with what was called “Distance Learning.” In those days, this involved taping courses and either playing them on dedicated (College) TV stations or providing the VHS tapes to students who then later completed assignments (either remotely or back at the school). I used to film those courses in one of those stations. This setup was considered quite revolutionary for the time. Not long after that I found myself teaching traditional courses in the classroom - giving me a view from the other side. Within 5 years or so though Distance Learning had already evolved to the next level both on computers and on the Internet. On computers Learning went through the CBT revolution (think CD ROMs) and then the first classes began appearing online. It was all happening so fast that in 1995 when I tried to get a Master’s thesis for “E-learning” approved at my college, it was rejected on the grounds that ‘there wasn’t enough scholarship (e.g. articles) on the topic published yet to pursue.’ Within 3.5 years of that, I was working on Cisco System’s E-learning Connection (after having taught computer courses onsite and developing CBT and online courses). By the end of the decade (of the ‘90’s) I even had a Yahoo Group called “E-learning Leaders.” In other words, I’ve been following this revolution for a long-time as both an enthusiast and sometimes participant.

In those formative years, I developed a vision for where I thought all of this was going to go; it involved greater levels of ‘intelligence’ and automation, but wasn’t thoroughly dependent upon it (at least not in the sense that ‘AI Dependency’ has now emerged). I later began to refer to that as “Dynamic Learning or “Me Learning,” given that the focus of the evolutionary path I foresaw was to build around individual empowerment. The idea in a nutshell was that the technology would allow both institutions and learners to become exponentially more flexible in how they managed both personal and collective Curricula. What has happened instead over the past two years or so with the widespread adoption of ChatGPT-like GenAI tools took me somewhat by surprise.

Instead of empowering learning itself, the GenAI tools are empowering Learners to bypass the learning process altogether and still gain acceptance within a pedagogical system that has remain largely unchanged. This is occurring across all conceivable learning scenarios, from K through Grad school. In other words, the AI platforms are not directly facilitating learning yet, but they are in many instances replacing it now.

This is a problem for a lot of reasons. Let’s cover the most important ones:

1. It has distracted from the needed reforms and evolution within the education system (which had it been more responsive to prior technologies may not have been so overwhelmed by the current ones). Education itself has not changed at all - and that’s why its so easy for AI to fool it.

2. It has already led to a new handicap; AI Dependency. This is a very real syndrome where younger people may literally not know how to deal with a wide variety of basic tasks because they never actually learned how to do them. There is a good analogy for this from previous technology. Reliance on Google Maps has conditioned a generation or two of Americans to not use real maps or learn how to find their way around without maps (e.g. to navigate). This is all the more ironic given that ‘way-finding’ has always been considered one of the toughest jobs for AI to master (and not navigating it has led to changes in human brain development).

3. It’s literally defeating the entire purpose of education. Having another entity think for you and do your work, (after having learned what you didn’t), is the exact opposite of empowerment. It in fact becomes a sort of slavery and seems on the face of it to be the ultimate dead end.

Where does this leave us?

When combined with the Great Replacement, which in fact this is very closely related to, AI’s redefinition of the educational experience will have a devastating impact on future generations if we continue going down our current path. This is something that needs to be addressed at a national and global level right now. So far, the reaction to AI in education has been twofold:

1. Tactically, teachers and institutions have been scrambling - somewhat unsuccessfully - to ‘catch’ students who are cheating with AI. This path is of course doomed to fail w/o more comprehensive structural changes given the rapid advance in AI capabilities.

2. Strategically, the powers that be have rushed to embrace AI - whatever that means. This hasn’t really been defined well in the educational realm outside the expectation that “AI Skills” will be reinforced in new curriculums. This of course will also fail miserably. Let’s look at why a little deeper.

There are no such things as “AI Skills”

And that’s kind of whole the point of the technology in question, which helps to highlight how out of touch most folks are with what’s happening. I’ve heard this term (AI Skills) thrown around constantly for the past two years and I find it remarkable that most people don’t seem to acknowledge that the only skill you need for these tools is the ability to write - and even that won’t be prerequisite much longer as voice prompts can and will work just as well. What this means is that there is no strategic options for how to deal with AI in the same way all previous technology trends were handled. Yet, you’ll continue to hear non-stop exhortations about how schools need to prepare students for the shining AI future by teaching these non-existent skills.

What do we do?

The first step towards solving a problem is acknowledging that you have one (this is the cardinal rule of consulting and perhaps of life as well). Everyone needs to understand (and quickly) that the current approach to Education will not survive the AI onslaught. Education itself is going to have to be reinvented because this situation is going to be with us from this point forward - forever. How do we reinvent Education in the Age of AI; here are some top-level suggestions:

• Stop focusing on technology skills; that’s no longer necessary and in fact will be a distraction from this point forward. Instead, core skills (such as analytical thinking) and preferred areas of interest should dominate.

• Eliminate homework in the traditional sense. Activities can still occur outside of the classroom, but all of those need to be ‘defended’ in the classroom to demonstrate that the students actually understand the material.

• (Finally) make Learning Learner-Centric. This is what I was talking about earlier when I mentioned “Me Learning.” Letting students pursue the things that interest them rather than just the boring standards-based material will make it far less likely that they’ll feel justified in cheating their way through. Motivation for learning is the ultimate super-power for improved Educational outcomes; we’ve ignored this for too long - now we’re being forced to face it.

I’ll be taking a closer look at the some of the AI Mitigations in Education in future articles here.

Copyright 2025, Digital Perspectives

I was wrong.

While we’re not facing the much vaunted “Singularity,” (the awakening of the Computer Consciousness that will then devour the world), we are in fact facing something else nearly as problematic - right now. As of 2024, something that many are referring to as the “Great Replacement” has begun. Like most situations, it isn’t and won’t happen overnight. But it is happening and the pace of it is beginning to quicken. The “Great Replacement” refers to the replacement of humans in the workplace by AI.

In a generic sense, this isn’t the first time that this has happened. The Industrial Revolution replaced age-old human roles and capabilities, then the Information Age came along and replaced still more and then back in the factories again, robots began replacing humans on the assembly line and in warehouses and now we are facing the combination of all these types of replacements (both White Collar and Blue Collar roles). All roads have led to automation and until just recently, all such innovations brought along with them at least some new opportunities to compensate for the loss of millions of human jobs and entire roles.

This time though, the new opportunities won’t be coming along with the innovations. AI itself will be the beneficiary of nearly every new skill or need associated with adoption of other AI automation (admittedly a bleak, yet realistic outlook). If there isn’t a deliberate movement put into place (soon) that will ensure that humans remain a significant part of various professions or job roles, we will be left behind. This is a problem only many levels; economic, political, cultural and of course on a personal level for all of us.

Now many folks are trying to claim that this isn’t happening; despite a lot of fairly persuasive evidence to the contrary; including:

• Elimination of roles such as Software Engineering where AI tools have already surpassed human capabilities.

• Weakening of the job market for new college grads (as lower-level roles are phased out through AI replacement).

• The vanishing of a number of job roles in near-real time - roles such as Customer Service Representative, Paralegal, etc.

• Massive Tech layoffs even in companies making handsome profits.

• The DOGE use of AI to replace Government roles (facilitating massive downsizing there of some of the best paying jobs in the country).

• Admissions from the AI Overlords in Silicon Valley that this is in fact happening and that they are actively working to make it happen.

While there is not now an never have been guarantees that there will be enough jobs in any economy to support its working age population, there was at least always the expectation that when times were better more opportunities would arise. But this expectation no longer applies in the Age of AI; if employment is only measured in terms of productivity and only AI automation can continue to accelerate that productivity at the rates desired by upper management (middle and lower management will no longer exist), then the outcome is certain. The workplace will evolve beyond humans.

How will we humans reckon with this?

This is the single biggest problem that isn’t be addressed by basically anyone. It’s been talked about a little; especially lately, but absolutely no one anywhere is doing anything about it. Perhaps, it’s assumed that there’s nothing to be done - that this Great Replacement will be the inevitable and unpleasant byproduct of progress and that we’ll all just have to deal with it. Or perhaps it’s a lack of imagination and courage at play here - with the AI moguls throwing around 100’s of billions of dollars like it’s Monopoly money and few if any politicians or governments willing to get steamrolled by it. It’s hard to tell why it’s flown under the radar this long, but the reality is that it can’t stay that way much longer as more and more people are displaced. Eventually, everyone is going to notice. Then what?

The Government & Industry Perspective

What can we expect from the the official channels in relation to this? Well, in a word - nothing. While there may be some movement in Europe, even there the reaction will likely be muted as the Tech Bros steamroll forward with mega data centers, nuclear power and a mission to replace as much of the workforce as fast as they possibly can - because that’s where the profits are (cutting out not just the middleman, but every man - labor is expensive you know).

The Cultural & Personal Perspective

This is the only realm where any reaction is possible and / or likely to happen. For most people, the personal sphere will involve the need to find new employment or perhaps entire new careers as their chosen fields are swallowed up. While this isn’t historically unusual, it will be harder to accomplish this time around and may require some fairly innovative (non-AI thinking) to succeed with it.

The best chances for success will likely be where individual action meshes with collective action on the Cultural level. This will take several forms:

1. The creation of a separate human-centric economy (that will compete with industries bereft of human inputs). For an analogy, think of the Organic Food industry. Many thought that such a thing was impossible several decades ago, yet it has taken off and become highly successful. This type of action will require a massive grass roots organization movement and it needs to start now.

2. Political Pushback. Human-centric work needs to become one of the major tenets of Economic Populism. Our current path is rapidly leading society to a hyper-concentration of wealth and the decimation of the Middle Class (globally). This type of action would necessarily involve pushback against all major parties as they are all more or less compromised by big tech dollars and will therefore be deaf to all of our concerns until they are forced to hear or them (or replaced by other parties). Again, this is something that needs to start now.

3. Some very specific thinking about what we wish (and or need to preserve) as “human domains.” For example, losing the ability to develop critical thinking and analytical skills, (due to over-reliance on AI in Education), will cause irreparable harm to society - in fact this is already happening right now. In order to stop this and to save other areas where humans - should - be involved, we need to identify what we’re going to save now. Think of it as like an Endangered Species Act, but for us. I took a shot at doing some of this identification last year, but hardly anyone else seems to be talking about it - and time is running out.

The Great Replacement is the real-time, real-world consequence of universal AI adoption and it’s here now - largely being ignored. Not good.

Copyright 2025, Stephen Lahanas