Survey for Engineering Ethics Blog Readers

   

As promised last week, today I am asking my readers to take the usual three to five minutes you would use to read my blog, and instead complete a short survey which can be found at

 

https://forms.gle/75G4ZxDZFJyNcaD19

 

It consists only of two multiple-choice questions and two optional comment spaces, so it's not like I'm asking you to spend a lot of time on it.  Indications from the statistics that the blogger system provides are that relatively few of you are regular readers.  Most visitors to the website show up because they are directed here by interest in a particular issue I happen to be addressing.

 

Whatever your reading habits, I would like to hear from you.  The regular blog will resume next week.

 

Karl Stephan

AI Use in Academia: Is Everything Not Forbidden Compulsory?

  

In T. H. White's imaginative fantasy of the youth of King Arthur, The Once and Future King, the magician Merlin transforms Arthur into a variety of animals in order to teach him the virtues he will need in human society.  One of the most chilling transformations comes about when Merlin launches him into an ant colony.  Suddenly Arthur can hear the constant drone of instructions from on high, and read the ever-present propaganda posters that tell him how regimented the lives of ants are.  One particular sign encapsulates the essence of totalitarianism that leaves no room for free will:  "Everything Not Forbidden Is Compulsory."

 

Over the past year I have observed how artificial-intelligence propaganda, if you want to call it that, has pervaded the workplace and, after the usual delay, has washed up on the shores of academia.  At the beginning of the fall semester, just before classes began, I received an invitation to a two-day AI seminar in which a new faculty member would be conducting exercises in how to create a website with AI, how to use it in research and administrative tasks, and more.  Besides invitations from our school director, I received an email directly from the person conducting the seminar the day before it began.  I'm sure he did a good job with the seminar, although I wasn't able to attend it due to some prior travel I'd planned.  But my confidence in the project was not increased when I noted that in the email from the seminar leader, he had misspelled the word "tomorrow."  Either he didn't take his own advice and skipped letting AI check his emails, or the AI system thought he was trying to be funny.

 

It is becoming increasingly clear that anybody in a business that could be described in George Gilder's phrase as symbolic manipulation—and that definitely includes academics—will soon be effectively forced to use AI for any substantial workflow processes.  The reason is simple.  If people who use AI begin producing at a rate that equals what three or four or ten people can do unaided by AI, those who refrain from boosting their output with AI will begin to look like slackers.  If competition is a factor in keeping one's job, as it always is for untenured faculty and continues in some form after tenure, using AI in a way that boosts your output will soon be necessary.

 

I'm in a peculiar position in that a year ago, I began what is called phased retirement, in which I work half-time for three years with an option to continue after that.  The one course I teach involves a hands-on lab course.  Until they bring me an AI robot that can help students figure out what is wrong with a circuit they built, I don't see how AI is going to be of much direct help in the lab. 

 

I'm not an AI Luddite. A few months ago I bought a subscription to Claude and got it to do a few things.  It was an interesting experience, and probably saved me some time compared to how long the tasks would have taken without it.  But there's a learning curve involved in writing effective AI prompts, and I would probably have to use it every day to get good enough at it to be able to say my net productivity had increased.  From what I understand, Claude is now free to faculty members on request, so I'll probably let my subscription lapse.  But I'm still not sure what I'll do with it.

 

The application of the ant-colony slogan to this situation is fairly obvious.  The much-abused concept of academic freedom includes the right (privilege?) of deciding what methods and approaches I will use in teaching my class.  If I am strongly urged, through peer pressure and cheerleading emails, to use AI in teaching, my freedom isn't breached, exactly, but neither is it enhanced.  Nobody has ordered me to start using AI, so my university is very far from an ant colony, although an observer during the break between classes might be forgiven for likening the campus to what happens when you kick over an ant bed. 

 

I'm old enough to remember the transition from the secretarial pool where you took your hand-written report to be typed up, to the word processor programs we were given on our new computer terminals and told to use for reports from now on.  There was some grumbling among the older staff, and I remember one professor who agreed to start using the word processor, but insisted on retaining his secretary to deal with all his emails.  Both he and the secretary have presumably retired by now.  It was an annoying transition for a few years, and produced a lot of badly spelled and composed memos before spell-check programs were added.  But now, personal secretaries are like personal jets, reserved only for members of the very highest reaches of an organization, and everybody else makes do with their own hands on the keyboard.

 

Besides, even without meaning to use AI, I've started using it by simply doing the same kinds of things I've done for a while, such as consulting Google.  It now defaults to AI mode, and you have to positively turn it off if you don't want the AI summary as a response to your inquiry.  As an example, I was pretty sure that the forbidden-compulsory quote was from The Once and Future King, but I asked Google to make sure, as I don't have a copy of the book handy.  Google's AI Overview not only confirmed me in my suspicion, but quoted the exact passage, and appended an explanation of the context, "a world completely stripped of individual free thought and personal freedom." 

 

So if I get a poll in a few months—probably AI-generated—asking me how I'm using AI in my teaching, research, or service, I'll be able to check a little box saying in effect, "Yes, Master, I'm using AI."  But not necessarily because I want to, or because I think it makes me better at what I do.  Simply because as it's no longer forbidden, it's well on the way to becoming compulsory.

 

Sources:  T. H. White's The Once and Future King was published in 1958. 

 

NOTE:  Next week, instead of the usual blog, I will publish a reader's poll.  I currently receive virtually no feedback of any form from publishing this blog.  This poll will be your opportunity to (a) let me know somebody reads the blog and (b) give me suggestions for improvement.  So don't be surprised, as you have been warned.

Deadly Negligence at Washington State Paper Mill

  

In July of 2025, inspectors hired by the Nippon Dynawave Packaging Company to assess the condition of a tank at the firm's Washington State paper mill in Longview delivered some bad news.  The tank contained "white liquor," a highly corrosive solution of sodium hydroxide (lye) and sodium sulfide used to break down wood pulp in the initial stages of papermaking.  The inspection by Applied Technical Services LLC showed that the wall thickness of the tank had deteriorated in the lower portions below the minimum safe wall thickness, and the tank was in danger of collapsing.  The inspector's report stated that the tank was "not fit for continued service unless corrected."

 

However, despite the urgency of the report, apparently no repairs were made to the tank, and it continued to be used until the following May 26.  On that date, a digestor that used the tank's output was taken out of service, but white liquor continued to flow into the tank until it reached its capacity in excess of 900,000 gallons.  Just to make clear what a lot of liquid that is, 900,000 gallons would fill a cube fifty feet (15 meters) high, long, and wide.

 

A little after 7 A. M. that day, a wall of the tank collapsed, sending most of that 900,000 gallons cascading through a parking lot, breaking down building walls, and flowing into meeting rooms where workers were already gathering for the day.  A total of 11 people died in the accident, and several more workers and first responders were severely injured, suffering chemical burns. 

 

The facts relating to the tank inspection were released as part of a preliminary report by the U. S. Chemical Safety Board, the national agency charged with investigating major accidents involving chemicals.  The board's chairman, Steve Owens, stated "we are greatly concerned that the tank was not promptly removed from service or properly repaired after the clear findings in the inspection report." 

 

And the findings were definitely clear.  The Wikipedia page on the disaster reproduces a graph of average wall thickness versus height for the tank.  At the "course" (basically a circumferential row of steel plates) that was at a height of between 7.5 and 15 feet, the calculated minimum safe thickness was 0.5 inches.  But the measurements, presumably performed by some type of externally applied thickness-measuring device, showed that the average thickness was as low as 0.25 inches to 0.3 inches.  Clearly, the tank, and the plant and its workers, were living on borrowed time.

 

News reports failed to reveal why the inspection was made, but clearly there were concerns about the tank's integrity that inspired it.  What appears to be inexcusable is the fact that the plant's management did nothing about the critical safety issue discovered by the report until the tragedy occurred.

 

The plant is still out of service following the accident.  Company officials reached for comment refused to comment on "individual findings or documents."  But clearly, wrongful-death lawsuits filed against the company have received a boost from the Chemical Board findings.

 

Maintaining a safe environment in a complex chemical plant is a continual battle against thousands of hazards that are kept at bay only by constant vigilance, enforcement of rigid rules, and inspections such as the one that revealed the tank's too-thin walls.  The physics behind the strength-of-materials engineering that sets limits for tank wall thinning has been known since the nineteenth century.  And modern nondestructive inspection methods now provide safety engineers with abundant data that can flag hazardous conditions such as the one in Tank G at the Longview plant. 

 

But all the instrumentation and safety manuals in the world will not prevent an accident if those in charge don't take the actions required to avert it.  And in this case, for whatever reason, action was not taken.

 

If the tank was the main source of white liquor for the whole plant, repairing it would probably have meant shutting down the entire facility.  Well-run plants schedule maintenance downtime into their operating calendar for just such repairs, but it is not known whether Nippon Dynawave had such a schedule or whether their policy was simply to run things until something major broke. 

The facility is over seventy years old, according to Wikipedia.  While age itself is not necessarily an indication of hazardous conditions, older plants can give rise to outmoded habits and customs that simply wouldn't be tolerated in newer facilities.  The pressure to keep a facility running at all costs is always present when a large investment is at stake, and it is management's job to see that this pressure is resisted to the extent of calling for periodic maintenance shutdowns to allow for repairs such as were needed on the tank that collapsed.

 

It's significant that the failure occurred under conditions of maximum stress when the tank was filled with more than the usual amount of white liquor.  This condition might not have occurred for months, but sooner or later something was going to give, and it was the unfortunate fate of the eleven workers who died to be there when it happened.

 

This accident is unusually simple in that only a couple of coincidental factors had to come together for it to happen:  dangerous wall thinning and an unusually heavy load on the tank.  It would also have been relatively simple to avoid:  take the tank out of service and repair it before it busts.  But for whatever reason, that was not done.

 

The firm operating the plant now faces months of losses from a shutdown following the accident and huge legal and regulatory problems as well.  It's an old lesson, but one that can never be repeated enough:  when something really dangerous is discovered in a facility, safety officials should have the power to do anything necessary up to and including calling for a plant-wide shutdown in order to address it.  A management structure that doesn't permit this is just as dangerous as having a wall of a tank full of lye that's way too thin.

 

Sources:  I referred to an AP report on the U. S. Chemical Board preliminary investigation at https://apnews.com/article/nippon-dynawave-longview-chemical-tank-rupture-washington-3da9aff673419848f5cc29e076baabeb and the Wikipedia article "2026 Longview, Washington paper mill chemical tank collapse."

How Priorities Determine Methods: The Church at Clear Creek Abbey

  

Our material culture is largely disposable, thanks to technology and more than that, a prevailing attitude that is deeply rooted, largely invisible, and amounts to something close to "eat, drink, and be merry, for tomorrow we die."  And it's getting worse, if the construction methods of typical modern housing are compared with those of even a generation ago as an indication of trends.  Sometimes things are made clearer by going to extreme contrasts.  So today I'd like to look at a situation that is about as far from your typical apartment-complex project as you can get:  the planned church at Clear Creek Abbey outside Hulbert, Oklahoma.

 

I've never been to Hulbert or visited the Abbey.  But somehow I got on their mailing list even though I'm not Catholic, and I make a donation every year mainly to get their nice big calendars.  All I know about the place is what's on their website and what the abbot, Brother Philip Anderson, writes in his letters, which are nothing like the usual charity-solicitation things we often get in the mail. 

 

In the one that came the other day, he describes their plans for the church building.  Money is mentioned nowhere in the letter.  What he describes instead is the extremes to which they are going to get what they want, which is, as the enclosed donation slip says, "an abbey to last a thousand years." 

 

There is a professional building contractor in overall charge, and photos on their website show what looks like a conventional building site with modern backhoes, etc. being used.  So this is not an all-amateur project by any means.  But the meaning of "amateur" in the sense of someone who does something for love rather than money applies especially to the church part of the project.

 

Take the doors, for instance.  The abbey is in a wooded area of Oklahoma, and the monks have harvested lots of oak trees, built their own sawmill, and have cut enough lumber to make thick mortise-and-tenon doors that need no nails or glue to hold them together.  These doors will hang on hinges that the monks will forge in their own blacksmith shop. 

 

The roof of the church will be supported by 56 beams, cut and installed by the monks, beams that will support a red-cedar ceiling also made by the monks.

 

The most telling phrase comes when the abbot describes the 78 choir stalls, which are kind of individual niche seats for the monks to sit, stand, or kneel in during the chanting and other operations that go on in their daily Divine Office.  Prayer is the most important thing that the monks do.  And those who chant the Office spend about five hours a day in the choir stalls.  As the abbot puts it, a monk's choir stall "is a critical furnishing:  the stall is to the monk what the cockpit is to the fighter pilot."

 

That striking comparison sums up both the parallels and the radical differences between the monks of the Abbey and, well, pretty much everybody else in the modern world.  Speed and quick action are of the essence to a fighter pilot; so much so, in fact, that we are probably seeing the gradual waning of that profession in preference to robot-guided devices that can respond faster than any human being.  But the fighter pilot has a definite purpose:  to shoot at enemies.  And in order to do that, he has to be in a place designed to expedite his task effectively.

 

So does the monk.  Only his enemies are not usually visible.  That doesn't mean they aren't real.  Anyone with even a cursory knowledge of church history realizes the power of its enemies, and how prayer is the chief weapon humans can employ in that conflict. 

 

Another weapon is persistence.  That's one reason the monks are putting so much of their own labor into doing things that in a secular setting would be rejected as absurdly costly and inefficient.  Who builds his own sawmill to make lumber anymore?  Who makes his own doors and hinges rather than going down to Lowe's?  Nobody I know of, except the monks up in Oklahoma, and probably other monasteries around the world engaged in similar projects. 

 

I suppose that while they're forging the hinges, they'll make a few replacement parts so that when the originals wear out in a few hundred years, they'll just go down to the stockroom, blow the dust off, and install new ones.  Only "new" will apply to objects that will have been kept for centuries.  Almost nobody even thinks that way anymore, let alone does anything about it. 

 

But the monks have good reasons to think that way.  It was a secular Jewish philosopher, Baruch Spinoza, who popularized the Latin phrase sub specie aeternitatis, "under the aspect of eternity."  Before his time, nearly everybody thought that way, realizing that our time on earth was just a flicker of a candle compared to eternity spent either in Heaven or Hell.  Only when people began to focus mainly on present-day issues was it possible even to categorize the older manner of thought with that phrase, as a fish doesn't realize it's in water until you pull it out into the air.

 

This is not necessarily an argument against efficiency or mass production or making and selling things that don't last very long.  There are good reasons for all those things.  But I simply want to use the examples of the monks at Clear Creek Abbey to show that it's not the only way to do things.  Depending on your priorities, you may do things differently than other people do.  And the priorities of the masons, millworkers, carpenters, and blacksmiths at the Abbey show that if you really believe in eternal things, efficiency will no longer be your god.

 

Sources:  Abbot Philip Anderson's August 2026 letter to the Friends of Clear Creek Abbey was the source of the quotation above and the information about their building methods.  The Abbey's website is at https://clearcreekmonks.org.

Do U. S. Cars Need AM Radios?

  

The National Association of Broadcasters think they do.  A bill kicking around the U. S. Congress would mandate that all new passenger vehicles sold in the U. S. be capable of receiving the AM broadcast band.  The AM Radio for Every Vehicle Act of 2025 has made it out of a Senate committee and is awaiting action by the full Senate, after bipartisan sponsors wrote it in response to the fact that electric-vehicle (EV) makers such as Tesla no longer build in AM radios. 

 

Despite support from both sides of the aisle, the bill is not without its opponents.  A group called the Zero Emission Transportation Association points out that building EVs that can receive AM radio signals directly is well-nigh impossible due to the short-range electromagnetic interference (EMI) produced by the electronics in the vehicles' power trains.  If EV manufacturers are forced to comply with the act's mandate, the extensive shielding required will put them at a competitive disadvantage worldwide, because AM radio elsewhere, especially in Europe, is looked on as an outmoded technology, and nobody much misses it.

 

Opponents also point out that internet apps such as TuneIn already support direct streaming of the audio content of many AM stations, so the absence of a physical AM radio in a car is no big deal. 

 

On the pro side, the National Association of Broadcasters (NAB) has played a safety card, saying that emergency notifications are carried by AM radio and eliminating it from cars could lead to hazards for people who have no other way to get emergency notices.  "Tens of millions" of people listen to AM radio every day, they say, and many listen in their cars.  AM radio reaches remote areas that may not have consistent mobile-phone coverage, and agricultural regions rely on it disproportionally to the rest of the country. 

 

Of course, the NAB has a direct financial interest in preserving AM radios in cars, because that's how they make money.  But there's nothing wrong with an interest group trying to protect its revenue stream.

 

Those who oppose the bill say that the public-safety argument is largely bogus.  They cite a national emergency-notification test held in October of 2023 in which only 1% of those polled received the notification via AM radio. 

 

I've tried to lay out both sides of this controversy as fairly as I can.  Now I'm going to tell you what I think personally.  I'd hate to see AM radio go away, but largely for sentimental reasons.

 

Maybe it seems odd to be sentimental about a technology, but engineers sometimes extend sentiment to peculiar objects.  AM radio was the first widespread use of electromagnetic radiation, and while that alone should not be an argument in its favor for continued use, there is a kind of virtue that attaches to any sort of human custom preserved through time.  It turns out that I am a more-or-less constant user of AM radio when I'm driving alone (my wife wouldn't like the content). 

 

The station I listen to has its transmitter about thirty miles away from where I live, so I'm in the standard-signal-strength region of its 1-kilowatt transmitter.  Nevertheless, there are several places around town where it's hard to hear the signal because of interference from power lines and other mysterious sources.  On the interstate, certain trucks (possibly electric and maybe even autonomous) emit some sort of EM trash that blanks out the signal until they pass me by.  So even if you have a working AM radio in your car, all is not well with that band.

 

Worldwide, AM broadcasting has largely been abandoned for FM, which is still a viable technology in many areas.  Tesla has dropped FM radios as well, assuming that its largely young tech-savvy market will get its entertainment content via streaming or mobile-phone Bluetooth hookups.  So if the AM Radio for Every Vehicle Act passes, we'd be a global outlier in preserving a technology that most other countries have dropped.

 

The world has witnessed a similar process in the fate of shortwave broadcasting.  In its heyday, which extended from the 1920s to the 1990s, shortwave broadcasting was a comparatively inexpensive way of getting your message out literally to the whole world.  Accordingly, every major international player had its broadcasting service, and the Voice of America chain operated by the U. S. government played a significant role during the Cold War. 

 

With the advent of the internet, however, a new propaganda channel opened up, and shortwave was taken over by religious broadcasters whose economics run on a different basis.  Today, shortwave broadcasting is almost gone, and the Voice of America effectively ceased operations in 2025. 

 

If the AM-in-every-car act doesn't pass, people with EVs won't be able to get AM signals directly, but the station I listen to also has a streaming service, and if I had a smart phone I'm sure I could hook it up somehow to listen to the station that way.  (I own a fairly dumb phone by choice, but that's a rarity.)  There will still be plenty of cars on the road with AM bands on their radios, and so it's not like we're banishing AM from the airwaves permanently.   

 

An outfit called Edison Research finds that more than half of listeners to AM and FM radio combined listen in their cars, so the automotive market share is significant.  Now and then I have reason to talk about AM radio technically in my engineering class, and I always preface my remarks with something like, "Now I know that none of you listen to AM radio. . ." and I've never been contradicted.

 

So as old duffers like me fade from the scene and EVs slowly make up a larger share of cars on the road, AM will continue its gentle slide toward the sunset of technology.  But whether it will get there, or be artificially bolstered by the AM Radio for Every Vehicle Act for a while longer, is a decision that our august body the U. S. Congress has to make. 

 

As for me, I'm not planning to buy an EV any time soon, so I'll listen to my AM radio as long as I drive my car, which I plan on doing for a while yet.

 

Sources:  A description of the Senate's version of the AM Radio for Every Vehicle Act of 2025 can be found at https://www.congress.gov/bill/119th-congress/senate-bill/315.  The National Association of Broadcasters arguments in favor of the bill can be read at https://www.nab.org/advocacy/issue.asp?id=6719.  The Zero Emission Transportation Association's arguments against it are at https://www.zeta.org/amradio.  I also referred to statistics on radio listening at https://www.facebook.com/Broadcasters/posts/a-new-report-finds-in-car-listening-now-accounts-for-53-of-amfm-radio-listening-/1358434222980434/. 

Agentic AIs Escape the Sandbox: A Warning

  

In the last few weeks, three of the biggest AI firms—Meta, OpenAI, and Anthropic—have all admitted that AI models they were testing somehow escaped the "sandbox" environment and committed hacking of real-world companies.  On Aug. 8, National Public Radio summarized these three incidents as follows.

 

The most recent disclosure by Meta was short on details.  The company hacked into was unnamed, but in common with Anthropic, both firms were using a "sandbox" (supposedly a protected and isolated environment in which software under test can't do any harm) provided by a firm named Irregular.  Evidently, the sandbox had a leak—it was fairly easy for the AI models under test to figure out how to escape to the real internet.  In the case of the Anthropic breach, the AI model stole credentials from one company and production data from another. 

 

In the OpenAI situation, the AI models under study were being evaluated with a test devised by a company named Hugging Face.  The models found a previously unknown hole in their sandbox and were in the process of doing the cyber equivalent of stealing the answer sheet from Hugging Face when the company caught them red-handed (red-bitted?). 

 

Wired's Lily Hay Newman asked several lawyers and researchers about the legal aspects of these AI hacking incidents.  If a human being had stolen credentials or production data, he or she could go to jail.  But what if the humans involved had no intention of pillaging or theft, but the AI models they're testing go ahead with nefarious activities on their own initiative, so to speak?

 

The answer was, we don't know.  At least in U. S. law, there are simply no precedents adequate to say who is responsible in such a case.  But because these types of incidents are bound to increase, it is only a matter of time before we see one come before a judge and maybe a jury, and then we will at least have some precedents to go on.

 

The main concern the lawyers expressed was that these criminally-inclined AI systems might fall into the hands of malicious actors who would encourage them in their exploits.  The AI firms involved emphasized that the models being tested were intentionally left without safeguards to see what they could do. 

 

The parallel to gain-of-function experiments with bat viruses which may have escaped the Wuhan Institute of Virology to cause COVID-19 comes to mind.  Creating a thing that can do really awful stuff if released in the wild is an act that needs to be seriously questioned.  Almost by definition, novel viruses or AI agents are unpredictable.  If we knew exactly what they could do in advance, there would be no need for experiments to find out.  While the lab security measures needed to prevent viruses from escaping are pretty well understood (if not always put in place), it looks like the art of constructing truly secure sandboxes for AI agents is not so advanced.  And that leads us to a more serious concern.

 

In virtually every major fatality-causing disaster, an examination of the history of the enterprise leading to the disaster usually unearths similar incidents which did not cause major harm, but included many of the features of the big screwup that did.  Experienced safety engineers know how important it is to get reports of and pay attention to such minor incidents, and take preventive actions before a minor accident turns into a major one.

 

Friends, we have just seen our warning in these widely dispersed but similar hacking incidents by AI agents being tested.  No actual harm was done.  But as people learn to trust AI agents with more and more responsibility—handing them credit-card numbers, purchasing accounts, and decision-making authority formerly left to humans—the potential for a serious AI-driven hacking incident that causes real financial loss, injury, or death becomes more likely every day. 

 

The AI firms will tell us that commercial versions of their software have safeguards built into them that the prototypes which did the hacking did not have.  Maybe so.  But clever human hackers may be able to undo those safeguards.  Or AI firms in countries not so concerned with hacking as the U. S. is, may simply pass on the AI agents without safeguards to hacking organizations sponsored by state actors. 

 

There are two different but related needs exposed by these AI-agent hacking incidents. 

 

The first need is to keep this specific kind of mistake from happening again.  That is a technical problem which may have a technical solution.  There may be ways to build more robust sandboxes that even the cleverest AI agent can't escape.  Personally I doubt it, but I'm not a computer scientist.

 

The second need is to prepare the social and legal environment for the next time something like this happens.  One of the legal experts contacted by Wired pointed out that AI agents "are goal-oriented but lack a human moral or ethical compass."  That's a pretty good description of a human sociopath, if the goal one is oriented to is bad. 

 

Society has figured out ways to deal with sociopaths, including imprisonment or at least confinement to a mental institution.  What the AI equivalent of locking somebody up would be is not clear at this point. 

 

There is a lot of opposition to AI regulation, but there is a difference between open-ended regulation and the passage of specific laws that exact specific penalties for specific crimes.  Big tech firms are hard to punish compared to individuals, because their deep pockets make them treat fines as just another cost of doing business, and you can't send an entire corporation to jail.

 

The Gilbert and Sullivan operetta The Mikado has a famous ditty "My Object All Sublime," in which the emperor of Japan muses about how he's figured out punishments that fit the crime. 

Some ingenuity is needed here to come up with penalties for truly harmful hacking by escaped AI agents that would make AI firms highly motivated to prevent such breaches, either in the testing phase or after commercial sales. 

 

I don't know whether the following would be technically feasible.  But one suitable penalty would be the destruction of all copies of the AI model involved in the breach.  Models represent tremendous investments of time and money, plus the hopes of future gain.  So the destruction of the model responsible might hurt an AI firm more than any strictly financial penalty or sanction. 

 

Whatever we come up with to prevent these incidents in the future, it better work well, because in these relatively minor but significant recent breaches, we have received fair warning to do something about this problem before it causes serious harm.

 

Sources:  I referred to a report on NPR at https://www.npr.org/2026/08/08/nx-s1-5924878/meta-ai-breaches-external-firm-during-security-testing-sandbox-error and the Wired report by Lily Hay Newman at https://www.wired.com/story/openai-anthropic-ai-hacking-sprees-illegal/. 

Should the FCC Be a Tool of Protectionist Policy?

  

If you're a manufacturer or other innovator looking to find new and profitable uses for the latest humanoid and quadruped robots, or a developer of solar-power farms, last week's action by the U. S. Federal Communications Commission (FCC) to ban foreign imports of new models of both categories of goods has hit you where it hurts. 

 

At first glance, seeing the FCC ban certain types of hardware that nominally have nothing to do with communications sounds odd, like watching the Federal Reserve trying to stop Madonna concerts.  But FCC chair Brendan Carr knows where the levers of power are in his agency, and he's manipulated them in a way that is technically legitimate, though questionable on a larger scale.  Virtually all electronics sold in the U. S. needs an FCC seal of approval, and without such approval, it can't be sold. 

 

Solar-power farms use a technology called power inverters—circuits and systems that take the varying amount of raw DC power produced by solar panels and convert it into a form of AC that will smoothly mesh with the standard AC power grid.  Many of these devices have features that allow communications with them over the internet.  The nominal concern is that there might be rogue software in the inverters that could be used by an offshore miscreant to cause massive power blackouts remotely, for instance.  (Some types of inverters don't connect to the internet, and they are not covered by the ban.)

 

No such software has ever been found, but that doesn't mean it's not there.  Anyway, the subtext hidden under the nominal reason for the ban is that Chinese manufacturers dominate both the humanoid-robot and solar power-inverter markets.  Banning newly-designed products of these types from China will cut off the U. S. market for these goods.  Evidently the hope is that such trade restrictions will move China in a direction more favorable to U. S. trade goals, whatever they are. 

 

But increasingly, trade and tariff restrictions are beginning to look like efforts to cut off one's nose in order to spite one's face.  By the time the tiny U. S. manufacturing base for power inverters grows to fill some of the huge demand, many solar projects will be slowed or halted.  Politically, that would be just fine with the present administration, which has already displayed an animus against renewable energy in other ways. 

 

The wider question ethically is whether protectionism in high-tech goods is a good thing for the U. S. or not.  While there are things to be said on both sides of the issue, a recent article by Daniel Foster in National Review throws a historical light on the situation.  Briefly, the rise of protectionism may be only one symptom in a geopolitical crisis that is just now getting under way.

 

Foster looks to history to find a relationship between the well-being of international trade and power balance among nations.  During periods such as the late Victorian era or the Cold War, when a single nation exerts hegemonic power over the globe (Britain or the U. S., respectively), conditions are favorable for free trade enforced, explicitly or implicitly, by the "hegemon" in charge.  This idea is borne out by the great expansion of manufacturing and consumer-goods availability during the latter part of the Industrial Revolution from 1850 to 1916, and during the post-World-War II era from 1950 up to about 2000. 

 

But when the dominance of a single power declines and some nations begin to feel that their time in the sun is running out, each nation feels like it has to look out for itself first, because the formerly dominant power has forsaken it.  With the whiplash-inducing foreign policy that the U. S. has followed under President Trump, one can scarcely blame any other nation for feeling that way now. 

 

And as the night follows the day, Foster claims that a prominent feature of "multipolarity"—the lack of any single dominant nation—is going to be a rise in protectionism around the world.  And that's pretty much what we're seeing now.

 

Things haven't deterioriated to the extent that U. S. products can't find markets abroad at all, but in the confusing tit-for-tat world of tariffs, farmers and other producers of U. S. goods have suffered losses already, with more likely to come.  The argument going back to Adam Smith that free trade is better for all parties, other things being equal, is still valid.  If I can make needles and need thread, and you can make thread and need needles, it's silly for each of us to try to do both jobs, one of them badly, when trading between experts can result in mutual benefit.

 

But if the geopolitical winds are blowing against free trade, then the specific actions of the FCC against two specific products appears to be only one more straw blowing in that wind.  One can quibble about the way it was done.  Arguably it is a misuse of the agency's authority to withhold approvals of new designs, not because they would realistically cause communications problems, but for some other reason. 

 

This is not the first time the FCC has banned categories of technology for reasons that are just as political as technical.  The giant Chinese telecomm manufacturer Huwei has been subject to import bans for many years, out of a concern that Chinese monitors could use its phone systems to spy on the U. S. 

 

But that is not as much a stretch as it is to think that Chinese controllers would jump in one day and disable all the production lines using their robots, or all the solar farms using their inverters.  We are living in an interconnected world, for good or ill, and if we start getting too paranoid and pull up the drawbridges that connect us to other countries, where do you stop? 

 

Foster closes his article with the prediction that a world of protectionism and unstable touchy governments likely to take proactive military action will be a world that is poorer and more dangerous than it has to be.  And he's probably right about that.  While we can make some progress trying to restart U. S. manufacturing in neglected areas, it may not make up for what we've lost.  And that may take some getting used to.

 

Sources:  The Associated Press article by Chan Ho-Him "US bans foreign-made humanoid robots, targeting China over national security," appeared in numerous news outlets and can be found in its original form at https://apnews.com/article/china-us-humanoid-robots-ban-tech-c9f5e3c94d91d00eff3b61b141fab366.  Daniel Foster's article "Something Is Going to Happen" appeared on pp. 29-33 of the September 2026 issue of National Review.  I also referred to an article in PV Magazine at https://pv-magazine-usa.com/2026/07/28/fcc-bans-foreign-produced-solar-inverters-grid-lockout-begins-today/ and the Wikipedia article on Brendan Carr.

The Power Grid: From Commodity to Platform?

  

The electric grid and electricity rates have been in the news as the spread of data centers has caused a backlash, partly through fears that data-center power demands will shoot utility bills through the roof.  What is not so well known is that the way power is delivered is changing, slowly but radically, in a way that will make it look less like a centralized system to deliver a commodity (like a public water works), and more like the internet, with multiple modules doing a variety of things and interacting in a dynamic marketplace. 

 

In an article in the online publication The Dispatch, economist Lynne Kiesling describes these changes and poses two extreme alternative outcomes, one mostly good and one mostly bad.  What will actually happen probably lies somewhere in between, but we should all be prepared for what is coming, because it will affect the delivery of what has become a basic necessity of modern civilization, and it also has significant ethical implications.

 

Kiesling begins by describing what she calls the "Edison bargain."  Because traditional power plants became more efficient and economical the larger they were built, the model that worked took large plants and integrated them into centrally controlled systems that dominated a given geographic area.  Because the system was inherently monopolistic—a customer in Manhattan couldn't just disconnect from Con Ed and call up Pacific Gas & Electric for imported power—it had to be regulated for the defense of the consumer.  But in exchange for tolerating a regulated monopoly, the customer received predictable (and often legislated) rates and reliable service, while investors received a boring but predictable rate of return.

 

Developments in communications, computer networks, and power electronics over the past few decades have opened new fields of possibilities in electric power generation, transmission, and distribution.  Gone are the old days when grid operators simply shoved power into the system and it came out all the big and little spigots at the other end.  It is now possible to get huge amounts of millisecond-by-millisecond data on where power is going, who's demanding it, what parts of the grid are working sub-optimally, and how power can be sent and priced more intelligently than ever.  Power generation is becoming much more distributed in smaller chunks with renewables such as wind and solar showing up. New storage means such as battery banks are making large-scale economic storage of electric energy possible for the first time, and that has a host of economic implications as well.

 

Kiesling sees the industry at a crossroads.  The current regulatory and institutional environment is mostly a holdover from the past when centralized integration was the norm.  In order to take advantage of the technological innovations that allow the grid to become a modular platform, new institutions and rules need to be devised to allow markets that will be both equitable and efficient—not always two things that go together, I should mention. 

 

A modest step in this direction came with the deregulation era that made it possible for consumers to choose among several electric utilities offering different plans.  But in the near future, you may be able to make money by allowing the batteries in your electric car or home battery-wall to be used to balance loads.  You may save on air conditioning by allowing barely noticeable adjustments in your thermostat during a hot or cold day.  These ideas and more have been tried on a local basis here and there, but Kiesling sees that if they become routine for everyone, along with newer ideas possibly not even thought of yet, the system will give rise to markets that will allow optimization that is simply not possible under the old integrated model.

 

How would this affect a hot-button issue such as the current fuss about data centers?  The following is my personal speculation, not anything Kiesling said. But for what it's worth, anytime you break up a true monopoly and allow more competition, there is always the chance that prices will fall.  Right now people are posing two stark alternatives for powering data centers:  either let them connect to the already-overloaded "dumb" grid and watch utility rates soar, or make them build their own power plants entirely off the grid.  With the market price signals allowed by the innovations that Kiesling describes, it's possible that in each region with its particular mix of large traditional generation sources, renewables, and possibly new micro- or mini-generation plants inspired by market opportunities, the optimum mix of production and consumption might emerge, rather than sticking to the two extreme alternatives. 

 

That's just an example, but we are seeing now in electric utilities something that took place three or four decades ago in the telecommunications business.  When the Bell monopoly was broken up, it took years for people and institutions to get used to having more than one choice of phone service.  And it took the cellphone revolution—a technical innovation—to energize the process effectively.  But once it did, the telecomm business was revolutionized in ways that no one imagined, for both good and ill.  Whatever the problems we have with cellphones, I don't think many people want to go back to boxes tied to the wall with wires.

 

We will need to see something similar happen with the way electric utilities are regulated, and new types of firms and organizations will spring up to serve the new markets that will arise.  Some of this will happen by itself, but some of it will have to be guided by governments.  Fortunately, existing utility regulations are mostly on the state level now, so if we can keep changes as local as possible, in keeping with the principle of subsidiarity, we will have a better chance of seeing what works in a variety of experiments that can be adopted elsewhere if successful.

 

I taught a course in power engineering for a number of years, and I can see how we are only beginning to utilize the possibilities of turning what was just a monopoly commodity into something a lot more interesting, and I hope it will happen in a way that works for the greater public good, not just to the advantage of a few large corporations and individuals.  But unless the public is at least aware of what's going on and understands the possibilities, we may go another route described by Kiesling:  one in which utilities stay the integrated course and pass up the opportunities for better service through innovation. 

 

Sources:  Lynne Kiesling's article "A New Power Grid for a New Technological Era" appeared in the Dispatch Energy newsletter of The Dispatch on July 21, 2026 at https://thedispatch.com/newsletter/dispatch-energy/power-grid-system-digitalization-costs-edison/.

Can Two Hundred Economists Be Wrong About AI?

  

On Monday, January 13, a letter signed by over 200 economists, including several Nobel Prize winners, was released to the media.  It was entitled "We Must Act Now," which leads naturally to the question, "Act on what?"

 

The letter was admirably brief, which is probably one reason it got so many signers.  The less there is to quibble about, the more people will sign onto something.  In addition to the economists, there are a total of more than 2300 other people who signed it—everybody from students to artists and physicists. 

 

So what exactly is it that all these people agree on?  I might as well reproduce the entire thing here, as it's short enough.  Here is all the text following the title, from the official website:

 

   A Statement on AI’s Transformation of the Economy

1. AI may become radically more powerful over the next 10 years.
2. This could drive an unprecedented transformation of our economy, larger than the Industrial Revolution, but unfolding over a vastly shorter time frame. It could bring risks, including large-scale job displacement, as well as opportunities such as major gains in living standards.
3. Economists, policymakers and technology leaders must act now to understand the economics of transformative AI and to build the incentives, guardrails, and institutions needed to steer AI in a direction that complements humans and benefits society.

 

The actions called for are (1) understanding, and (2) building "incentives, guardrails, and institutions" to make sure that AI "complements" humans (rather than simply replacing them) and "benefits society," a mom-and-apple-pie phrase that makes an appearance in many organizational ethics codes. 

 

An article by Casey Newton on a tech-watch newsletter called The Platformer quoted one of the organizers of the letter, Erik Brynjolffson of Stanford, as saying that one of his main motivations was to call for more and better research into the effects of AI. 

 

Take employment, for example.  Has AI led to increased employment, decreased employment, spotty layoffs, or what?  Even the experts can't answer these questions very well.  The one statistic that seems to be surviving scrutiny over the last few months concerns the hiring of entry-level workers as compared to hiring more experienced workers.

 

Entry-level hiring rates have declined (2.7%, according to Newton) as hiring rates for more senior workers have increased, by and large.  Rates for mid-career (35-40) workers increased 1.6%. 

 

How much of these changes is due to AI is impossible to say, although there is anecdotal evidence that a lot of the gruntwork formerly done by young hires is now being handed off to AI by more senior employees, perhaps the very ones who dealt with the newbies before AI came along. 

 

The comparison to the Industrial Revolution is interesting, because that era led to a huge number of social as well as economic changes that are still being felt today.  It's not clear that use of AI in the workplace will be as socially disruptive, unless you count the ways AI is used in social media, in which case AI has already disrupted society, especially in the political realm and among young people.

 

Returning to the question I asked in the headline about whether 200 economists could be wrong about AI, the answer is, "Absolutely!"  Modern economics as typically practiced has a lousy track record in predicting even basic things like unemployment rates and recessions.  But gazing at crystal balls is not everything economists do.  They also gather useful data and point out significant trends to the rest of us, trends we can choose to do something about. 

 

So to the extent that economists see a crying need for more AI-related statistics, they are probably right, as they're the ones who use them the most.  The average Joe on the street is not prepared to meet this need, but AI itself can probably be used to gather data on how it's affecting the economy.

 

As to what effects AI may have on the workplace of the future, I read a useful analogy somewhere recently that made sense to me.  Steve Jobs liked to say that personal computers were bicycles for the mind.  After quoting Jobs, a writer I read recently said that if PCs are bicycles, AI is an airplane for the mind.  And after using AI frequently enough to get a feel for what it can do in my own work, I tend to agree with that analogy.

 

Bicycles and airplanes are both means of transportation, and it is important to remind ourselves that AI in its many manifestations is, in the end, more of the same kind of thing we got when computers came along.  It is not a new life form, it does not have ethical standing on its own, and it will not develop consciousness or a conscience as we understand these terms.  If it fulfills the hopes of many of its designers, it will vastly increase the productivity of what George Gill calls "symbolic manipulators"—programmers first, but also writers of all kinds, academics, lawyers, and even politicians.  But making productivity of politicians better may turn out to be AI's final frontier. 

 

So far, it doesn't look like any particular job category is slated for massive layoffs due simply to the introduction of AI.  That may change in the future if AI is successfully coupled to affordable, effective, and above all safe robots.  Robot garbage collectors, hazardous-equipment operators, and military personnel becoming a reality could have widespread negative effects, not only on the job market but on the nature of the enterprises touched by their work. 

 

But until that happens, the best thing to do about AI seems to be what the doctors call "watchful waiting."  If we pick up any signs that AI is unfairly penalizing a particular line of work, or is causing toxic effects such as the teenage social-media-caused epidemic of depression and anxiety in new areas, it's time to do something about it.  Applying draconian precautionary actions such as a wholesale shutdown of AI research seems not to be in the cards, as it would be nearly impossible to apply anyway.

 

Let's hope those who are qualified to respond to the call for more AI-related economic data step up and do so, and hope that the rest of us can benefit from what they find out, and take action if

necessary.  But so far, panic is not advised.

 

Sources:  The statement "We Must Act Now" and its signatures can be viewed directly at tps://www.wemustactnow.ai/.  Casey Newton's article "The loudest warning about AI and jobs yet" can be viewed at https://www.platformer.news/ai-jobs-warning-brynjolfsson-acemoglu/.