Captain Obvious

Buzz Lightyear would be so proud of what the innovative thinkers are considering for the future location of digital server farms: orbiting our planet. The reasons stem from a host of issues here on earth that apparently become easier in stationary earth orbits. Take a look for yourself at: https://www.freethink.com/energy/future-of-data-centers and this commercial firm suggesting the same concept: https://lumenorbit.github.io/wp.pdf

Did you notice that no one seems to want to avoid this need in the first place? Can we really believe we can keep raping the planet making this stuff, killing children and their parents in third world countries as we irresponsibly mine because of our insatiable demand for raw materials, and then somehow find the power to fuel this unfettered growth forecast?

It seems like only yesterday that the electric utilities were promoting Redi Kilowatt and Gold Medallion Homes and forecasting load growth that forced them to build bigger and bigger power plants. The energy supply shortages in the 1970s changed all that and those plans were dashed on the rocks of energy efficiency, conservation, and least-cost planning.

The industry lurched toward natural gas to replace dirty coal, nuclear, and even hydro but ran into public opposition to virtually any form of new power plants. NIMBY (Not In My Backyard) was replaced with BANANA (Build Absolutely Nothing Anywhere Near Anybody) and NOPE (Not On Planet Earth).

How did we lose our memories of all this so quickly? Don’t we remember how we got here? Why are we on such a tear to build out capabilities that strip societies’ gears? There is a limit to sustainable life here on our planet and we are rapidly approaching it. Forget about the alleged climate crisis and look at the basic water, food, shelter, health care, and societal needs as we move much of the planet from third to second and then first world living conditions.

I know it is not politically acceptable to talk about restraining growth in our economy or standards of living, but we must start this dialogue and look creatively at longer term models for life. Otherwise, we will wind up with end points like the movies: WALL-E, Hunger Games, Idiocracy, and Waterworld.

Why do we listen to ridiculous answers like we will just go live on Mars or elsewhere?  Do we really think those ideas are better than working on fixing things here on earth? Or are we so elitist that we think we deserve a lifeboat for ourselves and just leave everyone else to forage for themselves?

An article by our good friend Dr. Kathrine Johnson (KJ) was the ultimate review of the realities of today’s AI Server Farm issues.  This seminal review offers an unvarnished assessment of the opportunities and challenges … so please read it here in its entirety.

Conundrums are basically intellectual puzzles with seemingly endless points of view and therefore no clear solution.  On one level, they can inform us about how complex something is … but they also disable decisions because you can’t see a clear path to success.   Let’s just face it … large corporations generally bog down in these situations.

The idea that nuclear power will somehow power these data centers is ludicrous at this time.  The time scale for approvals and building this type of generation is inconsistent with the decision timelines of those wanting to build data centers.  That will always be true, so get over it.  Move on … and stop thinking about wind and batteries please.

Those of you who want solar and batteries better look at the real estate needed to put up solar.  If you read KJ’s review you will see land is already contentious even for the server farm alone.

And, if you want to build any power plant, you must seek environmental permits.  They often take years and are far from certain because of many of the concerns raised by KJ in her article.  This is one of the biggest risks utilities have had when they need more units.

Therefore, the most likely way you will source your power requirements is by contracting with the regional electric utilities.  These are large loads, so if data centers are being located in Georgia, the advantage would be that any utility in the state could supply the power, using the electric grid for transport to the site.  But for most states, the grid will not support these large loads in rural areas.

The water-use challenge of cooling the power plant then moves to the serving utility, as do the environmental impacts.  But wait a minute … weren’t we worried about the carbon footprint and the consequential effects on our climate?  Where is that in this mix?

And you have the hype-cycle risk that most of this is just not real … what if the miracle being claimed is simply a mirage … another shiny object … and the market implodes?  I am far from alone in this concern, and if I am right, the financial outfall will be tragic … trillions of dollars wasted on another superficially appealing notion.

If this is just another boom-bust cycle, the winners will be analogous to those who sold blue jeans, shovels, and pickaxes to those attempting to mine the market, just like during the California gold rush.  These obviously include the AI modelling software companies who are in a frenetic race to outpromise each other and probably faking it till they make it … if they ever do.  Does Theranos come to mind?

And let’s take a closer look at one of the foundational assumptions with all of this as KJ’s article points out: the business model relies on these models helping with inference: a conclusion reached on the basis of evidence and reasoning.  So, the key here is whether the underlying data is adequately curated … does that remind you of my prior blog?

Well then, if we don’t have curated data, how can any of this produce sufficient economic value?  Plus, if we offer inferences from this kind of software, what bad consequences could result?  We can already see some of these risks with chatbots using AI:

It is a conundrum. I find the lack of humility around the consequences troubling.

 

The recent State of the Union address told me that our nation is in a very bad place.  The blatant lack of decorum and spirit for this country as a country was on full display.  It is now very clear to me, and perhaps to you, that our politicians view the situation as a zero-sum game.

For those of you not familiar with this term, it is the belief that one party’s gain is another’s loss. For instance, as in the game of chess.  One wins and the other loses; there is no other outcome other than perhaps the occasional stalemate. 

A recent Harvard study proved on average, Democrats were slightly more zero-sum than Republicans, with a greater tendency to view government as having a role in balancing inequities. But left-leaning voters with the strongest zero-sum tendencies also disproportionately split for Donald Trump in the past two presidential elections.

This recent working paper charts the surprising politics of zero-sum thinking — or the belief that one individual or group’s gain is another’s loss — with a goal of offering fresh insight into our nation’s schisms. The paper was co-authored by Stefanie Stantcheva, the Nathaniel Ropes Professor of Political Economy and founding director of Harvard’s Social Economics Lab.

Its analysis drew from detailed surveys of more than 20,000 Americans. It turns out zero-sum thinking does not map neatly with party affiliation. “But it certainly helps explain variations in people of the same political leaning,” Stantcheva said.

Source: “Zero-Sum Thinking and the Roots of U.S. Political Divides”

Vertical lines show the mean zero-sum index for each political party. “Republican” includes respondents who considered themselves “Strong Republican” or “Moderate Republican”, and “Democrat” includes respondents who considered themselves “Strong Democrat” or “Moderate Democrat.” Those who considered themselves “Independent” are not shown.

Informing the economics researchers’ approach was a rich body of previous research — including by anthropologist George Foster, the first to hypothesize in the 1960s that certain societies hold an “image of limited good,” with a firm belief in the finite nature of wealth and other resources.  Once again, the idea that the only outcome is that the winner limits the success of the loser by winning the game.

The result of this is the unending attacks and tireless talking-point banter between the parties, and the ostensible claim that eventually the parties will negotiate to some “middle ground” that frankly leaves both sides unhappy.

Let me offer the solution here: don’t seek middle ground … seek higher ground. Imagine you were God and living your life according to the greatest Old Testament commandment (Deuteronomy 4:6) and the wisdom of Micah in 6:8 to walk humbly and seek mercy.

What would you say to and about each other as you work towards workable near-term answers to the profoundly difficult challenges we face here and abroad?

Would you continue to throw stones at the adulteress, just because the law says you can and should?  Or, with the wisdom of the words attributed to Jesus, would you withhold your condemnation and possibly consider the greater good of redeeming the situation?

Are we teaching each other how to think this way, or are we just going to throw meat to our fan bases in the hope we will win this war of attrition? 

When will we learn that zero-sum game thinking never leads to peace on earth?  The temporary victors are always one season of politics away from retribution from those who lost in the previous round.

A recent review of Project Hyperion makes me wonder whether IQs have fallen lately.  I understand design competitions – they are often good ways to get intellectuals to think outside of the conventional boxes.  But this one just makes me wonder whether anyone on the design team truly understands sustainability and engineering fundamentals.

“Project Hyperion posed a challenge that most space agencies have sidestepped: how to design not just a vehicle but a self-sustaining human community capable of surviving a multi-decade journey between stars. The competition, run by the Initiative for Interstellar Studies (i4is) under the leadership of Dr. Andreas Hein, asked teams to produce blueprints for generation ships, vessels large enough to support entire populations from launch to arrival. The destination in most proposals was Proxima Centauri b, an exoplanet orbiting the closest star to our solar system and already a staple of speculative mission studies.

What set Hyperion apart from typical aerospace design exercises was its insistence on human factors. Entrants had to address governance, food production, psychological resilience, and reproductive planning alongside structural engineering. The competition’s published 2025 results reflected that priority: the winning entry did not promise the fastest engine or the lightest hull. It promised the most livable ship, treating social stability as a mission-critical system on par with life support or radiation shielding rather than an afterthought to be improvised once a vessel is already in flight.

So, Proxima Centauri b is about 4 light-years away … therefore, the journey time makes sense.  And the design teams rightfully worked through some of the logistical questions of reproduction and culture in a multigenerational experience like this.  That was all well and good.

But what was conspicuously missing was any sense of engineering, energy, and cost questions that this raise.  You have to get a vessel of this size built and moving at the speeds necessary to make the voyage and then solve the inverse problem as the vessel begins to approach its destination.  Plus, why would you do any of this to go somewhere to start a new planetary existence?

Star Wars fans would be thrilled with the 3D renderings and the seemingly feasible premise of doing this if money were unlimited and the raw materials for a project of this size were available.  The winning proposal for a 36-mile-long vessel makes people like me wonder how it could be built in near-orbit around our planet and then brought up to a realistic speed to achieve the stated goal.

The competition evaluators did have some concept of the problem:

“Most interstellar concepts lead with their engine, whether fusion drives, laser sails, or antimatter reactors. Chrysalis inverts that priority. The implicit argument is that propulsion technology will eventually catch up, but the harder problem, the one with fewer obvious solutions, is keeping a sealed human community functional across generations. Antarctic research stations, nuclear submarines, and long-duration spaceflight aboard the International Space Station all offer partial analogs, but none approach the scale or permanence that a generation ship demands.”

Yeah … and if you can solve these problems for Project Hyperion, why wouldn’t it be better to just build model communities here on earth now to make life here sustainable?

Why aren’t we doing the math?  Oh … I keep forgetting … 97% of the world’s scientists agree with what their funders tell them is a problem.  The real problem here is that we are not asking the right questions in the first place.

Educators are in a tizzy, just like back when calculators first entered the classroom.  The need to memorize the multiplication tables has been marginalized.  I noticed that my colleagues in engineering could no longer estimate solutions because they no longer used a slide rule.  For those who never used one, it only gives you the first 2 or possibly 3 digits to the answer … you have to know where the decimal point goes.

Now that AI is essentially free and widely available, people are asking it all kinds of questions and, unfortunately, just assuming the answers are reliably correct.  They fail to recognize that AI is simply parroting what the public sources offer as consensus opinions, and these opinions depend upon how you phrase the question.

I have warned all my friends to ask questions with opposing implied points of view.  For example, instead of asking a seemingly simple question: Is Jesus the only way to God? Ask the more general one: “what do the major world religions view as necessary beliefs to find peace with their concepts of God.”

I hope you can see the foundational difference between a curious person and one who is simply looking for the talking points to support their point of view.  With all this as context, please consider this Wall Street Journal article on how AI helps or thwarts teamwork:

https://www.wsj.com/opinion/how-i-killedand-revivedteamwork-with-ai-5a5722cf?st=mMMFLX&reflink=desktopwebshare_permalink

Now, back the blog title question.  Does AI make us lazy?  The answer is no.  We can be lazy as we go through life, and AI can make that addictively easy.  Or we can decide to be curious and helpful to ourselves and our communities by becoming more aware of the diversity of opinions and perspectives around us.

We don’t have to, and shouldn’t, decide we are right and they are wrong.  The tensions between these potential conclusions should serve as a basis for a humble search for wisdom and wellness in our relationships.

It is tempting, and all too easy, to use AI to just build bigger walls around our worldviews, smugly concluding that others are simply less well-informed or stupid in our minds.

Calculators enabled us to do more difficult math more accurately and more quickly than ever before.  Failure to think critically before we enter numbers into these calculators remains our responsibility if we want the best answers.

Go back to the basics of all human interactions:  people don’t make you mad … what you believe about them and the situation is what you are allowing to make you mad.  They probably have no idea what you are thinking because they, too, are just being lazy about your interaction.

Break the bad habit of being lazy … be curious … ask questions … listen … and ponder.  It’s OK to say a situation is complicated and you are trying to understand all perspectives.  Stop being intellectually lazy … AI doesn’t do that to you.