Artificial Intelligence: This Month’s Fine Progress, Real and Potential

From March 6^th to March 20^th, there was a string of articles – two about areas of true AI achievement, one questioning what may or may not have happened earlier this year, and two trying to persuade us about its ability to reach certain lofty heights. 

In the first, we saw how “Taco Bell shows off AI ‘coach’ following massive digital tech investment” (Pilar Arias, Fox Business, March 6^th).  Already, “about 500 Taco Bell U.S. locations have AI voice technology to take drive-thru orders,” up from 100 nine months before, and its Byte software, at least partially in use at 25,000 Taco Bell and KFC restaurants, “is widely scaled and enables operational consistency and restaurant manager efficiency,” including “online and mobile app ordering, point of sale, kitchen and delivery optimization, menu management, inventory and labor management, and team member tools.”  We will look for information no later than next year on how well that is working out.

The other told us, about a patient, that “A.I. Saved His Life by Discovering New Uses for Old Drugs” (Kate Morgan, The New York Times, March 20^th).  The wife of a man “battling a rare blood disorder” and looking at near-certain death contacted a doctor they had met, who used an AI model to propose “an unconventional combination of chemotherapy, immunotherapy and steroids previously untested.”  The patient was “responding to treatment” within a week, and within months was healthy enough for the “stem cell transplant” he needed to put him into remission.  In other cases as well the technology can provide “a systematic way” of assessing “a treasure trove of medicine that could be used for so many other diseases,” which a Harvard Medical School professor called an “enticing alternative” for treating rare afflictions.  Other successes include an AI proposal to use isopropyl alcohol inhalation for nausea which “worked instantly,” ADHD-dedicated amphetamines to help “periodic paralysis in children with a rare genetic disorder,” a drug used for Parkinson’s helping “patients with a neurological condition” to “move and speak,” and more.  This immensely valuable sort of thing, assimilating existing information on side effects, which in these cases were desirable, and other medication properties, more voluminously than humans could do, may be a perfect task for AI.

“How long will the DeepSeek euphoria last?” (Don Weinland, The Economist, March 15^th).  The author said that in that case “many China watchers are now trying to figure out what is real and what is froth,” as “it is worth keeping in mind that China is prone to asset bubbles,” and tension between China and the US could mean that “some of the green shoots might be short-lived.”  We still don’t have the well-established truth about DeepSeek’s asserted ability to produce AI modules at a small fraction of what other companies have paid, and until we do, any comments about its future are premature.

The first tall peak was the subject of “Lila Sciences Uses A.I. to Turbocharge Scientific Discovery” (Steve Lohr, The New York Times, March 10^th).  Per Lohr, “the big, inspiring A.I. opportunity on the horizon, experts agree, lies in accelerating and transforming scientific discovery and development.”  Well, there are others, one of which we will rediscover in the next paragraph.  The Cambridge, Massachusetts startup in the title “had worked in secret for two years to build scientific superintelligence to solve humankind’s greatest challenges.”  It has “already… generated novel antibodies to fight disease and (has) developed new materials for capturing carbon from the atmosphere,” and “turned those experiments into physical results in its lab within months, a process that most likely would take years with conventional research.”  As a result, “many scientists” think “that A.I. will soon make the hypothesis-experiment-test cycle faster than ever before,” and it “could even exceed the human imagination with inventions.”  Although “the early projects are still a long way from market-ready products,” the company “will work with partners to commercialize the ideas emerging from its lab.”  So there is plenty more to do, and while seeing reason for real optimism, comprehensive success is still an unknown amount of time away.

Even more optimistic was New York Times technology columnist Kevin Roose, as he showed in that publication’s “Why I’m Feeling the A.G.I.,” published on March 14^th and in the Sunday, March 16^th print edition.  He wrote “I believe that very soon – probably in 2026 or 2027, but possibly as soon as this year – one or more A.I. companies will claim they’ve created an artificial general intelligence, or A.G.I., which is usually defined as something like “a general-purpose A.I. system that can do almost all cognitive tasks a human can do.””  That means “we are losing our monopoly on human-level intelligence, and transitioning to a world with very powerful A.I. systems in it.”  “Powerful A.I.” will also “generate trillions of dollars in economic value and tilt the balance of political and military power toward the nations to control it.”  Reasons he names for his position are that “the insiders are alarmed,” “the A.I. models keep getting better,” and “overpreparing is better than underpreparing.”  This viewpoint harkens back to two years ago, before most people came to see AI’s doubts, impediments, and shortcomings more clearly.  There have been recent calls for those producing the technology to focus more on specific applications, some as strong as those described in the first two pieces here, and less on the hope of AGI, and this piece does not cancel them out.  Therefore, we have controversy between experts, and the rest of us don’t know either.  When we get much closer, then, I will keep you posted.  In the meantime, this may have been the best all-around month for AI since February 2023.

Autonomous Vehicles, Starting and Stopping – Where Are They Going Now?

We’re way past any idea that driverless cars will soon become the norm – but they are still progressing, in the taxi niche and elsewhere.  What have they shown us in the past 3½ months?

Where one large automaker’s products won’t be going is to give rides for hire, as “G.M. Will Stop Developing Self-Driving Taxis” (Jack Ewing and Eli Tan, The New York Times, December 10^th). They will still work on autonomous vehicles, as General Motors “said it would fold its Cruise subsidiary, which was working on that project, into its main operations, allowing formerly separate development teams to jointly develop fully autonomous vehicles for private owners.”  Although “the decision removes G.M. from a business that some in the industry believe could someday be worth hundreds of billions of dollars,” the company’s CEO “suggested that the payoff was too far in the future to justify the expense,” and said it “would focus on technology that will allow vehicles sold to consumers to steer, accelerate and brake without driver intervention under certain conditions.” 

Recent mishaps include “Passenger in Waymo self-driving car gets stuck circling parking lot while trying to make flight” (Pilar Arias, Fox Business, January 7^th).  The man called Waymo customer support, who fixed the problem in “just over 5 minutes” whereupon the vehicle delivered him to the airport.  The piece did not say if, as the passenger asked, the car had been hacked.

In further expansion, we have “Uber offering driverless rides in major Texas city” (Daniella Genovese, Fox Business, March 4^th).  Uber can offer Austin customers “a Waymo fully autonomous all-electric Jaguar I-PACE vehicle, which will allow them to travel across 37 miles in the area.”  Those two companies “joined forces in 2022 to bring autonomous rides to the public.”  To get such service, “riders till first have to opt in” to it, and those “who request an UberX, Uber Green, Uber Comfort or Uber Comfort Electric could be matched” with such a vehicle, which they can choose to accept.  “When the car arrives, riders will be able to unlock the vehicle, open the trunk, and start the trip in their app.”  It will cost the same as the usual amounts for these four Uber variations, and, in case you were wondering, users “won’t be prompted to tip.”

As well, a “Fleet of Amazon-backed self-driving taxis will soon hit the Las Vegas streets with public rides” (Sunny Tsai, again Fox Business, March 6^th).  “Ten years in the making,” they will be provided by Zoox, and “don’t look like standard cars,” as “they have four seats inside the vehicle, and riders face each other, sitting in rows of two.”  They “will be able to call their taxi on an app,” and “rides will be competitively priced to other existing offers.”  All of that could become a standard for additional service expansions.

Most recently, we merged automata, driverless vehicles, and freight capacity, as a “New delivery robot can haul 2,200 pounds of your stuff” (Kurt Knutsson, Fox News, March 9^th).  It looks like a truck with two rear ends, with its “194 cubic feet of cargo space” designed to be “modular,” so “it can be customized to suit various business needs.”  Although the article used the present tense throughout, it did not mention areas or specific locations where these are in use.  Let’s hope it really is, and that the progress shown in the last four articles here – I include the second, since the problem resolution time was so short – represents what we will continue to see.

Artificial Intelligence: Three Months, Six Different Facets

Since mid-December, except for the DeepSeek brouhaha AI has been relatively quiet.  There have been small expansions here and there, and sales of chips for it have stayed strong.  It has neither crapped out nor exploded.  Here are some looks at various views of it.

First, though, I address something my regular readers may have wondered about.  Why have I not been posting on the actions of the Trump administration, many important to American jobs and the American economy?  The answer is simple: they are changing too quickly.  We have gone back and forth on tariffs, job cuts, and foreign policy so swiftly that recounting those actions, and their reversals, is not suited to a weekly blog.  I sense that President Donald Trump is using them as bargaining tools more than for true policy changes, so the actions of other countries and other organizations, along with his choices of responses, will determine what will become permanent, what is temporary but briefly implemented, what is temporary without execution, and what turn out to be only threats.  For various reasons, I do not think the quantity and intensity of these edicts will continue for anywhere near a four-year presidential term, so we will be able to assess their significance, when the roller coaster slows down or stops, sooner.  In the meantime, I recommend reading the New York Times, if you want to keep track of the daily situation.

Back to the topic, the first piece here is “Business leaders commiserate over scaling roadblocks at AI Summit” (Patrick Kulp, Emerging Tech Brew, December 12^th).  You might think that a conference three months ago would be ancient AI history, but no, it doesn’t move that fast.  The “scaling” was more about selling, “failure to launch,” or a need to “fully commercialize” AI, as two conference presenters put it.  Progress can be slower than some may have thought, as well; for example, it may seem conceptually easy to book airline flights, but that “is still four or five years off.” 

“Why Does OpenAI Need So Much Money?”  This query, placed by Cade Metz on December 17^th in the New York Times, is in response to the company consuming $10 billion in the 18 months starting early 2023, and expecting to do the same with $10.6 billion ending in mid-2026.  The reasons are needing “enormous amounts of data” calling for “larger and larger amounts of computing power from giant data centers,” thereby using chips and electricity which are scarce in the quantities it requires.  And it may get even more expensive, as OpenAI “chases the dream of artificial general intelligence” (AGI), potentially “a machine that can do as much as the human brain, or more.”

Progress announced by a certain Chinese firm has not shrunk these numbers, and even if it was honestly represented, it may never, as “DeepSeek Doesn’t Scare OpenAI, Thanks to the ‘Jevons Paradox’” (Talmon Joseph Smith, The New York Times, February 14^th).  That holds “that as a resource becomes more efficient to use, demand will increase.”  It dates from 1865, when it applied to coal, and has a “decent track record” since then.  As we will see, not all pressures on AI are pushing it up, but this one has real potential to do that.

What does it mean when the “Microsoft CEO Admits That AI Is Generating Basically No Value” (Victor Tangermann, Futurism, February 22^nd)?  The executive, Satya Nadella, “has had it with the constant hype surrounding AI,” and, per the author, “argued that we should be looking at whether AI is generating real-world value instead of mindlessly running after fantastical ideas like AGI.”  The CEO also asked that “when we say this is like the Industrial Revolution, let’s have that Industrial Revolution type of growth.”  As “OpenAI’s top AI agent… still moves at a snail’s pace and requires constant supervision,” Nadella’s statement “could be seen as a way for Microsoft to temper some sky-high expectations.”  Yet he, himself, agreed during the previous month to invest $80 billion in the Stargate industry project.  So maybe, as bizarre as that sounds, it doesn’t mean much at all.

Per a Goldman Sachs-connected survey, “A majority of small businesses are using artificial intelligence” (Kennedy Hayes, Fox Business, February 27^th).  The majority found was 69%, and most found it was saving both time and money.  I do not know the minimum size of these ventures. 

Yet one business application has its enemies, as shown in “Interviewing on AI:  Tech companies love AI.  Just don’t try it to use it to get a job at one” (Dan DeFrancesco, Business Insider, February 28^th).  Amazon has been whining that “AI tools give candidates an “unfair advantage,” and that it doesn’t allow Amazon to evaluate their “authentic” skills, according to “guidelines” given to its internal recruiters.  Too bad, so sad.  Sorry if that is harsh, but companies, especially those of Amazon’s caliber, use plenty of proprietary technological methods to assess, consider, and reject candidates.  Fair is fair for jobseekers to use software of their own.  More power to them – and more power to everyone understanding, comprehending, and assessing the true merits of artificial intelligence.

February’s Employment Data Didn’t Change Much, With 151,000 Net New Jobs and AJSN Showing Latent Demand Almost Stationary at 17.1 Million

This morning’s Bureau of Labor Statistics Employment Situation Summary wasn’t the important one – that will be next month, when it will have had time to absorb the current governmental chaos.  But it still gives a good look at how employment was doing before the effects of the back-and-forth tariffs and employment cuts impacted the data.

We added 151,000 net new nonfarm payroll positions, only a bit short of the 160,000 and 170,000 estimates.  Seasonally adjusted and unadjusted unemployment each gained 0.1% to come in at 4.1% and 4.5%.  The other numbers tracked here got worse:  the adjusted count of unemployed gained 300,000 to 7.1 million, those out for 27 weeks or longer increased 100,000 to 1.5 million, the employment-population ratio and the labor force participation rate each sank 0.2% to 59.9% and 62.4%, and those working part-time for economic reasons, or keeping such jobs while thus far unsuccessfully seeking full-time ones, lurched up 400,000 – is there a story there? – to 4.9 million.  Average private nonfarm hourly payroll wages rose 6 cents, less than inflation, reaching $35.93. 

The American Job Shortage Number or AJSN, the measure showing how many new positions could be quickly filled if all knew they would be easy to get, sat in nearly the same place, gaining 27,000, as follows:

The most-changed contribution to the AJSN was from those discouraged, adding 150,000 less, almost offset by more people unemployed and more reporting they wanted work but did not look for it over the past year.  The share of the AJSN from those unemployed was up 0.5% to 39.8%.  Over the previous year, the AJSN rose less than 100,000, with a large boost in those officially jobless more than equalized by a smaller number of American expatriates.

What does all this add up to?  The number of those out of the labor force decreased 140,000, and those not interested fell 300,000, meaning that there was no great worker-pool inflow or outflow from existing people.  While the overall tone of the report was down, we can’t take those 151,000 new opportunities for granted.  Accordingly, the turtle didn’t go forward, but didn’t backtrack either.

On the Edge to a New Paradigm for Human Behavior – I

One of the more amazing books I have ever read crossed my desk earlier this winter. 

It is Nate Silver’s 2024 On the Edge: The Art of Risking Everything.  The dust jacket notes say it “investigates “the River,” the community of like-minded people whose mastery of risk allows them to shape – and dominate – much of modern life.”  That group is made up of “professional risk-takers” who are “poker players and hedge fund managers, crypto true believers and blue-chip art collectors.” With “high tolerance for risk, appreciation of uncertainty, affinity for numbers,” and “an instinctive distrust of conventional wisdom and a competitive drive so intense it can border on irrational,” they “can teach us much about navigating the uncertainty of the twenty-first century.”  The book explores the mindset of River members, through the remarkably interrelated areas of poker, sports betting, gambling in general, statistics below the surface, Las Vegas as contrasted with other American cities, general principles of risk-taking, artificial intelligence, the search for situations with positive expected value (ones expected to have value over their costs in the long run), venture capital theory and practice, a number of philosophical perspectives, and, above all, how to quantify whether and to what extent an action is worthwhile.  It runs 559 pages, including 24 on a glossary of “how to speak Riverian,” or definitions of about 400 concepts Silver used elsewhere in the book, which might be its most valuable part.  It is complicated but remarkably readable.  It is destined, unless Silver quickly outdoes it, to be a classic, in business, psychology, philosophy, economics, game theory, and probably in other subjects.

As with any great work of art, it also lends itself to different interpretations and branches of thinking not inherent to the original.  One has occurred to me while thinking about, re-reading (and re-re-reading, re-re-re-reading…) sections of it, and applying it to my life.  Here is my spinoff.

The book has a dichotomy, between those Silver classified as being in the River, and others in its rival and in many ways opposite community, the Village, which is “reflected most clearly in intellectual occupations with progressive politics, such as the media, academia, and government (especially when a Democrat is in the White House).”  The two communities “consist overwhelmingly of “elites,”” and “the vast majority of the population doesn’t fall into either group.”  Silver saw problems with each, but clearly was himself a Riverian, so his perception of the value of Village thinking may be too low. 

My different way of looking at people, though, while related to Silver’s scheme, parts company from it importantly and includes most of the people left out.  That I call being “Silver,” for the author, or “Lead,” for the base metal.

What is the difference between Silver and Lead, and how does that vary from this book?  In general, Silver ideas, actions, and general people are not only courageous but prudent about risk-taking.  Unlike the “degens” (out-of-control heavy gamblers) and people who take risks at greatest expense to others (e.g. Donald Trump), Silvers weigh those factors, and make sure, when taking a risk, not only does it have truly positive expected value (or enough recreational value to justify it), but it is sufficiently undamaging to bystanders.  Maybe even more than in the examples in this book, they identify areas, which can be recreational, social, sexual, or non-financially enriching, where the downside is minimal but the potential upside is large.  That they do through assessing the true disadvantages, taking their distrust of conventional wisdom far enough to question the often-perceived inappropriateness of certain actions.  For example, a man may be considered “creepy” for asking a woman he does not know, in whom he has romantic interest, if she is married, whereas if he stops interaction the moment it clearly becomes unwelcome, it may be perfectly inoffensive.

Here is a starting list of attributes and people that generally fit in the two categories.  Silver behavior usually or always includes courage, risk-taking, initiative-taking, fantasy fulfillment, independent thinking, saying yes, social openness, trying to involve others, more sex, pushing the envelope in general, remembering “you only live once” (or twice, for your dreams as well?), reevaluating old reasons for not taking action, preferring to beg forgiveness instead of asking permission, much rule-breaking, intensity, vitality in general, most extroversion, being internally honest about what they want and don’t want and don’t care about, questioning whether limitations other people said they have are valid, self-discipline, thinking outside the box (especially way outside the box), or doing things their way.  Lead thinking gets emphasis on protocol, risk-averseness, unwillingness to defy unnecessary-seeming conventions, being other-driven, being stopped by unjustified fear, being only a spectator, watching large amounts of TV, and the opposites of the Silver items.  In general, Silvers soar and let you soar; Leads pull you down.

I also started a list of people in my life who were Silvers and Leads, some of whom were Silvers at some times and Leads at others.  I found it enlightening, especially when considering those I liked or did not like.  If you are interested in this material, think about those you have known – it may be instructive for you as well.

Much more to follow.

Unions These Days – Recent Events, and What Their Presence Indicates

Trade unions have a long world history, mostly but hardly exclusively beneficial.  Now it is 2025, and few alive can even remember the pre-union times when, for example, if American workers died on the job they would not even get that day’s pay.  Labor laws, assuring physical and legal employee protections are extensive, widespread, noncontroversial and effective, and few companies run afoul of them.  So where do unions come in now?

First, a few of the large, related stories from the past nine months.

In “The Delivery Business Shows Why Unions Are Struggling to Expand” (The New York Times, May 27^th), Peter Eavis contrasted worker-organization victories, as those “representing workers at three large automakers and UPS negotiated new labor contracts that included big raises and other gains,” with “labor experts” saying “structural forces would make it hard for labor groups to increase their membership.”  One of those is the large share of workers not being employees of the companies at whose locations they report.  Others are increasing part-time work, high turnover in fields otherwise ripe for organizing, and frequent requirements that unions start simultaneously at all of an employer’s often-far-flung locations.  Still, unions have done some amazing things, exemplified by the statistic that the “average annual compensation, including benefits” of UPS drivers is now $170,000. 

Moving to efforts at another huge, well-known company, Matt Bruenig wrote in the August 21^st New York Times that he saw “In a Union Triumph, the Seeds of Future Failure.”  Starbucks employees, he noted, “have unionized 481 stores with more than 11,000 employees in less than three years,” which has, though, also revealed that “American labor laws, and the bureaucracy they require, make mass unionization impossible unless rules for certifying unions and negotiating contracts are simplified and streamlined.”  The National Labor Relations Board, responsible for holding many union elections as well as dealing with worker complaints, has been overtaxed.  “Anti-union activity by employers” is still a problem, and it is not even the law that union authorization requires only most workers signing certification cards.  While bills with legislation ending these concerns have been written and presented, they not been successfully voted in – and seem unlikely to be over the next several years.

The situation at the coffee seller reached a head four months later, as “Starbucks Baristas Walk Out in 3 Cities” (Heather Haddon, The Wall Street Journal, December 21^st -22^nd, 2024).  They were Chicago, Los Angeles, and Seattle, and the picketing, “walkouts,” and “protests” were over five days later, without agreement on a contract.  Later, we saw “Starbucks and Union Agree to Mediation in Quest for Contract” (Danielle Kaye and Rebecca Davis O’Brien, The New York Times, January 30^th), as the company, which “called the union’s wage proposals “not sustainable,”” “did not offer a substantial wage increase during the latest bargaining session in December.” Starbucks’ management has choices to make, on which they will be forced if they do not resolve them freely.

Also last month, and with a group of employees rather better compensated than baristas, there was a “Port Strike Averted With Labor Deal Days Before Deadline” (Peter Eavis, The New York Times, January 8^th).  The dockworkers’ union, the International Longshoremen’s Association, and the United States Maritime Alliance representing employers, “overcame their differences over a big sticking point in their talks:  the introduction of automated cargo-moving machinery at the ports.”  The resolution cemented an agreement on wages to go up over 60% during the time between now and 2031, and calls for positions to “be added when automated equipment was added at a port,” which will give hirers “a more straightforward path for introducing automated machinery.”  By the end of the six years, dockworkers will be getting $63 per hour, and “with shift work and overtime, the pay of many longshoremen at some East Coast ports could rise to well over $200,000 a year.” 

With container ship commerce vital nationally, the International Longshoremen’s Association is in an exceptionally strong bargaining position.  But that is not the case for unions at Starbucks or Amazon.  They are new, and the reason for their ascent is clear: their constituents have legitimate gripes.  That’s what emerging unions need now.  With truly inhumane treatment of employees almost completely a thing of the past, it is time for companies to take fresh union activity as a wake-up call.  What are they doing wrong?  It’s something.  They need to be aware of that before the organizing starts.  If they do, it won’t happen.  If they don’t, and end up in Starbucks’ position of needing to pay more than they think their businesses can take, they have only themselves to blame.

Artificial Intelligence Progress, Problems, and Perceptions: Two Months’ Worth

Recently, the DeepSeek kerfuffle has dominated the AI news.  But since late December, other things beyond future-success assertions have hit the news.  What are they?

First, “OpenAI Unveils New A.I. That Can ‘Reason’ Through Math and Science Problems” (Cade Metz, The New York Times, December 20^th).  Its new product, o3, now in the hands of “safety and security testers, outperformed the industry’s leading A.I. technologies on standardized benchmark tests that rate skills in math, science, coding and logic,” being over 20% more accurate than its predecessor “in a series of common programming tasks.”  That area has received less publicity, but AI has been serving more and more in human programmers’ core roles.  Yet, per OpenAI CEO Sam Altman, “at least one OpenAI programmer could still beat the system on this test,” meaning AI has not taken over superiority yet, and it “can still get things wrong or hallucinate.”  But we’ve seen a big improvement here.

The upcoming main OpenAI system, though, does not look as good, as “The Next Great Leap in AI Is Behind Schedule and Crazy Expensive” (The Wall Street Journal, December 21^st).  Author Deepa Seetharaman said about GPT-5, code named Orion, that “it isn’t clear when – or if – it’ll work,” as “there may not be enough data in the world to make it smart enough.”  This product, intended to succeed GPT-4 and its variants, is being developed with what could soon be regarded as the old way of building AI, with ever-more-gigantic datasets and similarly huge amounts of electricity and processing power, in this case costing “around half a billion dollars in computing costs alone.”  It’s now almost two years in the making, and still does not have even a traditionally overoptimistic release date.

Moving to an application, we saw “Platonic Romances and A.I. Clones: 2025 Dating Predictions” (Gina Cherelus, The New York Times, January 3^rd).  The author predicted less conventional dating and less use of conventional dating applications.  AI may become “your ultimate wingman” as it may be more frequently used to “write… profiles, edit photos and write entire dialogues… on dating apps,” and “some will even use A.I. clones to do the whole thing for them” as people have done with structurally-similar job applications.  As well, people may “use A.I. dating coaches to practice chats before a date, help them come up with conversation topics and suggest preplanned date ideas in their cities.”  At that point, AI had better be able to produce unique output streams, since it wouldn’t be much fun to be able to anticipate what a prospective romantic partner is next going to say, word-for-word.

Among our president’s immediate proclamations was “Trump announces largest AI infrastructure project ‘in history’ involving Softbank, OpenAI and Oracle” (Brock Dumas, Fox Business, January 21^st).  Despite those company’s CEOs joining “Trump from the Roosevelt Room at the White House for the announcement,” there is doubt whether they will meet the cost of “$100 billion, with plans to expand to $500 billion over the next four years” – and if they do, the project may serve only as a framework for capital expenditures they were planning to make anyway.

Another jump in actually-available AI capability occurred with “OpenAI launches Operator – an agent that can use a computer for you” (Will Douglas, MIT Technology Review, January 23^rd).  The software “can carry out simple online tasks in a browser, such as booking concert tickets or filling an online grocery order.”  There are already similar tools at Anthropic (Claude 3.5 Sonnet) and Google DeepMind (Mariner).  This one may be a cause for security worry, as there are plenty of ways of initiating physical action from a keyboard, so these apps will need to be constrained somehow.

“When A.I. Passes This Test, Look Out” (Kevin Roose, The New York Times, January 23^rd).  It has been produced by the Center for AI Safety and Scale AI, and it is called “”Humanity’s Last Exam.””  It has “roughly 3,000 multiple-choice and short answer questions designed to test A.I. systems’ abilities in areas ranging from analytic philosophy to rocket engineering,” which are “”along the upper range of what one might see in a graduate exam.””  If researchers can sufficiently restrict sharing these questions and their answers, this exam might last a while without AI solution, or it could topple within a year or two.

 A new look at an old subject, “Is Artificial Intelligence Really Worth the Hype?” was written by Jeff Sommer and published February 7^th in the New York Times.  Its main cause for concern was about DeepSeek, and, as of this date, investors were “re-evaluating prominent companies swept up in A.I. fever, including Nvidia, Meta, Alphabet, Microsoft, Amazon, Tesla and the private start-up OpenAI.”  In the week since this piece came out, there have been serious concerns voiced about the legitimacy and credibility of DeepSeek’s claim, and, since there is no other focus of unease in this article, we can’t accept it yet.  There are still plentiful reasons to think AI will not even approximate its highest expectations, and there will be more stories with similar titles, but the DeepSeek controversy – and that’s what it is – has not been resolved yet.  Neither has a clear vision of what AI will be doing, and not doing, even a few years from now.  We’ll see – that’s all.