Claims that can’t be checked by Gary Marcus, page 5 of 24
Deadlines are judged as of 2026-10-29. About Gary Marcus and the sources covered.
Claims that can’t be checked by Gary Marcus (2355)
Public evidence cannot settle these claims, so they are recorded and attributed, not scored. Most were screened out: both AI providers read the claim on its own and found it a value judgment, rhetoric, hedged, too vague to test, or about something no public evidence could show. The rest were found to be recommendations or too open to test in a full review.
Page 5 of 24: claims 401 to 500 of 2355, ordered by the date first said, newest first. To find a claim by its words, use search.
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“which it itself built on and extending an earlier (May 2025) dissection from Futurism that The New York Times should have read and addressed in far more detail:”
Our reading: The New York Times should have read and addressed in far more detail Futurism's May 2025 critical dissection of Medvi before praising it.
No horizon
Source: The back story behind the first “$1.8 Billion” dollar “AI Company”
Assessment: Can’t be checkedSaid
Observation
“With actual AGI nowhere in sight, redefinition and spin may be all that’s left.”
Our reading: Actual AGI is nowhere in sight as of April 2026.
No horizon
Assessment: Can’t be checkedSaid
Causal claim
“A quarter-billion dollar desperation move: an effort to control the narrative, to distract away from the billion or so dollars they are losing each month this year, and rumors that their stock is losing favor on the secondary market.”
Our reading: Marcus's only guess is that OpenAI's $250 million purchase of TBPN is a desperation move: an effort to control the narrative and to distract from the billion or so dollars it is losing each month this year and from rumors that its stock is losing favor on the secondary market.
No horizon
Qualifier: his only guess
Assessment: Can’t be checkedSaid
Rhetoric
“If I am not mistaken, I guess makes cell phones and Microsoft Word … “superintelligent”.”
Our reading: Marcus suggests, tentatively ('if I am not mistaken, I guess'), that Suleyman's definition would make cell phones and Microsoft Word 'superintelligent'.
No horizon
Qualifier: if I am not mistaken
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“I see at least three reasons off the top of my head, all of which should have been obvious before the bombs were dropped.”
Our reading: The reasons AI is poor at war strategy should have been obvious before the US-Iran war began.
No horizon
Qualifier: before the bombs were dropped
Source: In the Iran war, it looks like AI helped with operations, not strategy
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“AI is fine for writing memos; nobody should rely on it to plan a war or to guess its likely outcomes.”
Our reading: Nobody should rely on AI to plan a war or to predict its likely outcomes.
No horizon
Source: In the Iran war, it looks like AI helped with operations, not strategy
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“In the meantime, the advice is simple: Don’t focus on replacing humans. Focus on how you can use AI to help the ones you’ve got.”
Our reading: For now, employers should not focus on replacing humans with AI and should instead use AI to help their existing workers.
No horizon
Qualifier: in the meantime
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“There is a parallel lesson for employees, of course. Don’t panic, at least not yet.”
Our reading: Employees should not panic yet about AI replacing their jobs.
No horizon
Qualifier: at least not yet
Assessment: Can’t be checkedSaid
Conditional forecast
“All this could change; probably someday it will—but most likely not until we see more radical advances in AI, which could be a decade or more away.”
Our reading: The so-far modest impact of AI on productivity and return on AI investment could change, and probably someday will, but most likely not until there are more radical advances in AI, which could be a decade or more away.
No horizon
Only if: not until more radical advances in AI
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If your humanoid robot can’t understand the visual world, it’s just a demo, and not something you can trust.”
Our reading: A humanoid robot that can't understand the visual world is just a demo and not something you can trust.
No horizon
Scope: humanoid robots that can't understand the visual world
Source: The mirage of visual understanding in current frontier models
Assessment: Can’t be checkedSaid
Forecast
“Also, re the very active discussion on AI and jobs: although some white collar jobs (e.g., entry-level coder or market research assistant) may be in near-term jeopardy, many of those that require visual understanding (architect, cartographer, civil engineer, film editor, medical illustrator, urban planner, etc) probably aren’t vulnerable until entirely new techniques are developed.”
Our reading: Some white collar jobs, such as entry-level coder or market research assistant, may be in jeopardy from AI in the near term.
No horizon
Qualifier: e.g., entry-level coder or market research assistant
Source: The mirage of visual understanding in current frontier models
Assessment: Can’t be checkedSaid
Conditional forecast
“Also, re the very active discussion on AI and jobs: although some white collar jobs (e.g., entry-level coder or market research assistant) may be in near-term jeopardy, many of those that require visual understanding (architect, cartographer, civil engineer, film editor, medical illustrator, urban planner, etc) probably aren’t vulnerable until entirely new techniques are developed.”
Our reading: Many jobs requiring visual understanding (architect, cartographer, civil engineer, film editor, medical illustrator, urban planner) probably won't be vulnerable to AI until entirely new techniques are developed.
No horizon
Only if: until entirely new techniques are developed
Source: The mirage of visual understanding in current frontier models
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The bad bargain of AI data centers makes no sense.”
Our reading: The bad bargain of AI data centers makes no sense.
No horizon
Source: BREAKING: Sam Altman concedes that we need major breakthroughs beyond mere scaling to get to AGI
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“With the case for scaling as a road to AGI steadily crumbling, it is time to reconsider.”
Our reading: Because the case for scaling as a road to AGI is falling apart, the plans to spend heavily on data centers should be reconsidered.
No horizon
Qualifier: case for scaling as a road to AGI crumbling
Source: BREAKING: Sam Altman concedes that we need major breakthroughs beyond mere scaling to get to AGI
Assessment: Can’t be checkedSaid
Observation
“Taken together with Musk’s recent admission that xAI was “not built right” and Zuckerberg’s delay of Meta’s latest model, the shift from three prominent tech CEOs in a short period of time is a strong sign that insiders are losing faith in pure scaling.”
Our reading: The recent shifts by Altman, Musk and Zuckerberg strongly suggest that AI industry insiders are losing faith in pure scaling.
No horizon
Source: BREAKING: Sam Altman concedes that we need major breakthroughs beyond mere scaling to get to AGI
Assessment: Can’t be checkedSaid
Rhetoric
“(Dario Amodei’s ridiculous claims about doubling life span in the next decade are deluded on this front.)”
Our reading: Dario Amodei's claims that AI will double human life span in the next decade are deluded technosolutionism.
No horizon
Qualifier: in the next decade
Source: F Cancer
Assessment: Can’t be checkedSaid
Forecast
“Of course AI can have a positive impact on medical research long before it reaches that level of sophistication, simply by streamlining everyday tasks, helping with subject recruitment and paperwork, and so on; such productivity gains seem plausible in the short-term, even if the more dramatic magic some people are hoping for doesn’t.”
Our reading: In the short term, AI will plausibly bring productivity gains to medical research through mundane tasks, even if more dramatic breakthroughs do not arrive.
No horizon
Qualifier: via streamlining everyday tasks, recruitment, paperwork
Source: F Cancer
Assessment: Can’t be checkedSaid
Observation
“The one thing I think Javorsky gets somewhat wrong is that she downplays just how important getting the right algorithm is, and how far we are from that point.”
Our reading: Getting the right AI algorithm matters a great deal for beating disease, and current AI is far from that point.
No horizon
Source: F Cancer
Assessment: Can’t be checkedSaid
Forecast
“In the near-term, the greatest contributions from AI towards combating disease will probably come from enhancing humans with better tools rather than radical new scientific discovery.”
Our reading: In the near term, AI's biggest contributions to fighting disease will probably come from giving humans better tools, not from radical new scientific discoveries.
No horizon
Source: F Cancer
Assessment: Can’t be checkedSaid
Causal claim
“AI will also need to sweat the details a whole lot better.”
Our reading: Actually beating cancer will also require AI to sweat the details a whole lot better.
No horizon
Scope: for actually beating cancer
Source: F Cancer
Assessment: Can’t be checkedSaid
Rhetoric
“So much for pure scaling”
Our reading: Marcus dismisses pure scaling in light of the Meta and xAI results ('so much for pure scaling').
No horizon
Source: BREAKING: Expensive new evidence that scaling is not all you need
Assessment: Can’t be checkedSaid
Observation
“That hypothesis continues to go badly, as expensive experiments from two of the world’s wealthiest men have just shown.”
Our reading: The hypothesis that spending more on compute and data is the road to AGI continues to go badly, as recent expensive experiments by two of the world's wealthiest men (at Meta and xAI) have just shown.
No horizon
Qualifier: as shown by recent expensive experiments at Meta and xAI
Source: BREAKING: Expensive new evidence that scaling is not all you need
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If you actually believed that Claude was sentient and that sentience was de facto a supply chain risk, you would have to think that essentially every LLM is a supply chain risk. In which case the right choice is to stop using LLMs altogether, not to selectively punish Anthropic.”
Our reading: Anyone who actually believed that Claude was sentient and that sentience was de facto a supply chain risk would have to think that essentially every LLM is a supply chain risk, in which case the right choice would be to stop using LLMs altogether rather than selectively punish Anthropic.
No horizon
Only if: if you actually believed Claude was sentient and that sentience was de facto a supply chain risk
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Claude is next-word predictor, not an authority on its own internal states. It’s silly to take its estimates seriously.”
Our reading: Claude is a next-word predictor and not an authority on its own internal states, so it is silly to take its estimates of its own consciousness seriously.
No horizon
Scope: its estimates of its own probability of being conscious
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If you think it is bad to use hallucinating LLMs to autonomously choose targets (I certainly do), you shoudn’t use any of them.”
Our reading: No hallucinating LLM should be used to choose military targets autonomously.
No horizon
Qualifier: for autonomously choosing targets
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Which is to say we shouldn’t be putting our probabilities before our definitions, or even putting specific probabilities on things we can’t define.”
Our reading: People should not put specific probabilities on AI model consciousness before defining what consciousness means.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If you are worried about systems having guardrails, you shouldn’t be using LLMs, because you certainly shouldn’t be using them without guardrails.”
Our reading: LLMs should never be used without guardrails, so anyone worried about guardrails should not use LLMs at all.
No horizon
Qualifier: if worried about systems having guardrails
Assessment: Can’t be checkedSaid
Observation
“But there is no more reason to think that Claude is sentient than to think that Google’s now long-obsolete system LAMDA was.”
Our reading: There is no more reason to think Claude is sentient than to think Google's LaMDA was.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“They mimic things humans say, but we shouldn’t take them as veridical descriptions of anything happening on the inside.”
Our reading: LLMs mimic human speech, so their statements should not be treated as accurate descriptions of their internal states.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“We may well move to a regime in which AI writes most code — but for a long time to come we are going to need humans to fix the mess.”
Our reading: AI may well come to write most code.
No horizon
Source: “A spate of outages, including incidents tied to the use of AI coding tools”, right on schedule
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“As I wrote yesterday, Dario Amodei is no saint, but I fully support his company’s new lawsuit against the US government.”
Our reading: Gary Marcus fully supports Anthropic's new lawsuit against the US government, although he says Dario Amodei is no saint.
No horizon
Qualifier: Dario Amodei is no saint
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“I stand with them. Everyone should.”
Our reading: Everyone should stand with Anthropic in its lawsuit against the US government.
No horizon
Qualifier: in its lawsuit against the US government
Assessment: Can’t be checkedSaid
Observation
“Claude is (or at least was, and may again be) part of the war machine; even with humans putatively in the loop, it may have been responsible for many deaths.”
Our reading: Claude is, or at least was and may again be, part of the war machine, and even with humans putatively in the loop it may have been responsible for many deaths.
No horizon
Qualifier: even with humans putatively in the loop
Assessment: Can’t be checkedSaid
Observation
“Wright may well be right that Anthropic’s Claude selected the school; it would be hardly surprising if accidents like these happened.”
Our reading: Claude may well have selected the Iranian elementary school that was struck.
No horizon
Assessment: Can’t be checkedSaid
Causal claim
“The thing about including humans in the loop is that if the AI is selecting 80 targets an hour, as has been reported, the humans probably aren’t going to do a good job verifying those targets.”
Our reading: If the AI is selecting 80 targets an hour, as has been reported, the humans in the loop probably aren't going to do a good job verifying those targets.
No horizon
Only if: if the AI is selecting 80 targets an hour
Assessment: Can’t be checkedSaid
Observation
“In many other ways, Amodei pretty much follows Sam Altman’s playbook, of constant hype and overpromising, frequently moving deadlines back and rarely taking responsibility.”
Our reading: In many other ways, Amodei pretty much follows Altman's playbook of constant hype and overpromising, frequently moving deadlines back and rarely taking responsibility.
No horizon Evidence as of
Limitation: This does not endorse or reject Marcus's characterization.
Assessment: Can’t be checkedSaid
Rhetoric
“But aside from his courageous stand on mass surveillance and fully autonomous weapons, Dario is more like Sam than almost anyone I can think of.”
Our reading: Aside from his courageous stand on mass surveillance and fully autonomous weapons, Amodei is more like Altman than almost anyone Marcus can think of.
No horizon
Qualifier: aside from his courageous stand on mass surveillance and fully autonomous weapons
Assessment: Can’t be checkedSaid
Observation
“We have seen it before in law (hallucinated cases that wind up in submitted briefs), journalism, and so on. Now we are (likely) seeing it in war.”
Our reading: Overtrust in AI, already seen in law and journalism, is likely now being seen in war.
No horizon
Assessment: Can’t be checkedSaid
Rhetoric
“Amodei’s claims are ludicrous, at best. And plain dishonest, at worst.”
Our reading: Amodei's claims about AI are ludicrous at best and dishonest at worst.
No horizon
Assessment: Can’t be checkedSaid
Observation
“In Iran, I am going to suppose that humans were in the loop (about which I will say more in a moment) but even so the results probably already have been disastrous.”
Our reading: The results of AI-assisted targeting in Iran have probably already been disastrous, even supposing humans were in the loop.
No horizon
Qualifier: even with humans in the loop
Assessment: Can’t be checkedSaid
Rhetoric
“As bad as Sam Altman is, Dario Amodei isn’t all that different.”
Our reading: Dario Amodei is not all that different from Sam Altman.
No horizon
Assessment: Can’t be checkedSaid
Rhetoric
“In many ways, Amodei and Anthropic are simply speedrunning Altman’s fall from grace.”
Our reading: Amodei and Anthropic are rapidly following Altman's fall from grace.
No horizon
Assessment: Can’t be checkedSaid
Rhetoric
“History will judge those who stay at his company.”
Our reading: History will judge negatively those who stay at OpenAI.
No horizon
Source: BREAKING: Sam Altman’s greed and dishonesty are finally catching up to him
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Anyone who wants to work on LLMs can work elsewhere. Anyone who wants to use LLMs should go elsewhere.”
Our reading: People who want to use LLMs should use providers other than OpenAI, and those who want to work on LLMs can work elsewhere.
No horizon
Source: BREAKING: Sam Altman’s greed and dishonesty are finally catching up to him
Assessment: Can’t be checkedSaid
Rhetoric
“And all this was entirely predictable. Altman is bad news.”
Our reading: The backlash against Altman was entirely predictable; Altman is bad news.
No horizon
Source: BREAKING: Sam Altman’s greed and dishonesty are finally catching up to him
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“And, now at long last, the world sees what the board saw, and what I saw (and what Karen Hao saw): having someone running a company with that much power to affect the world who is not consistently candid is not a good idea.”
Our reading: A person who is not consistently candid, such as Altman, should not be running a company as powerful as OpenAI.
No horizon
Source: BREAKING: Sam Altman’s greed and dishonesty are finally catching up to him
Assessment: Can’t be checkedSaid
Observation
“In reality, Altman was never really all that interested in AI for the “benefit of humanity.” Mostly he was interested in Sam. And money, and deals.”
Our reading: Altman was never really all that interested in AI for the benefit of humanity; mostly he was interested in himself, and in money and deals.
No horizon
Qualifier: never really all that interested
Source: BREAKING: Sam Altman’s greed and dishonesty are finally catching up to him
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Given all that, the second part of the title of this essay (“don’t trust it with people’s lives”) should be obvious.”
Our reading: Generative AI should not be trusted with people's lives, given that it cannot even be trusted to do taxes.
No horizon
Qualifier: given that it cannot be trusted to do taxes reliably
Source: Don’t trust Generative AI to do your taxes — and don’t trust it with people’s lives
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Wanna feel good about yourself? Use a chatbot. Want to find truth? Go elsewhere.”
Our reading: People who want to feel good about themselves can use a chatbot, but people who want to find truth should go elsewhere than chatbots.
No horizon
Scope: for those who want to find truth
Source: Breaking: “sycophantic AI distorts belief, manufacturing certainty where there should be doubt”
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“But they have also been influential, and not in a good way.”
Our reading: Doomer AGI projections have been influential, and not in a good way.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“History will probably regard that as a turning point.2”
Our reading: History will probably regard the Pentagon's rejection of Anthropic and contract with OpenAI as a turning point.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Now we all may wind up paying the consequences. I am fully with the doomers in wanting to prepare for the future, but that needs to start with a realistic assessment of the present.”
Our reading: Preparing for the AI future should begin with a realistic assessment of current AI.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“And in so doing perhaps lead to thousands or millions of unnecessary deaths.”
Our reading: Governments' use of unreliable AI in war, by possibly leading to the accidental acceleration of war, may perhaps lead to thousands or millions of unnecessary deaths.
No horizon
Qualifier: via accidental acceleration of war
Assessment: Can’t be checkedSaid
Forecast
“Which (precisely because it is unreliable) may lead to the accidental acceleration of war.”
Our reading: Government use of unreliable AI in war may accidentally accelerate war.
No horizon
Qualifier: because the tech is unreliable
Assessment: Can’t be checkedSaid
Conditional forecast
“And we may wind up with a boy-who-cried wolf situation when true AGI does arrive.”
Our reading: When true AGI arrives, warnings may be ignored due to earlier false alarms.
No horizon
Only if: when true AGI does arrive
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“And it should be incumbent on them to understand the limitations and inaccuracies of the algorithms they choose.”
Our reading: Those who use military AI should be obliged to understand its limitations and inaccuracies.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“In my own view the biggest problem with the algorithms targeting Gaza was not necessarily the algorithms per se (about which not much may be public) but the decision to tolerate a large number of civilian casualties as part of the targeting.”
Our reading: In Marcus's own view, the biggest problem with the algorithms targeting Gaza was not necessarily the algorithms themselves, about which not much may be public, but the decision to tolerate a large number of civilian casualties as part of the targeting.
No horizon
Qualifier: in my own view
Assessment: Can’t be checkedSaid
Conditional forecast
“Meanwhile, unless and until the military does actual empirical studies on collateral damage, we won’t really know whether AI is helping or hurting.”
Our reading: Without empirical military studies on collateral damage, we won't know whether AI is helping or hurting in targeting.
No horizon Evidence as of
Only if: unless the military does empirical studies on collateral damage
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“(More broadly, the military use of AI needs to be a granular question.”
Our reading: Military use of AI should be evaluated granularly, task by task.
No horizon
Assessment: Can’t be checkedSaid
Observation
“Altman, for his part, doesn’t seem to care, having signed a contract entirely full of holes, despite making noises about red lines.”
Our reading: Altman signed a military contract full of holes despite talking about red lines, and seems not to care.
No horizon
Assessment: Can’t be checkedSaid
Causal claim
“Mileage may vary according to the task, and may be worse in unfamiliar situations, given the inherent tendencies of generative AI.)”
Our reading: Because of its inherent tendencies, generative AI's military performance may vary by task and may be worse in unfamiliar situations.
No horizon
Qualifier: military tasks
Assessment: Can’t be checkedSaid
Forecast
“Some may lead to further escalation (a mass killing of school children could well do that); in the worst case, a series of escalations triggered by AI-triggered mistakes could lead to a nuclear war.”
Our reading: AI-triggered military mistakes may lead to escalation and, in the worst case, could lead to nuclear war.
No horizon
Qualifier: worst case
Assessment: Can’t be checkedSaid
Forecast
“We may get little detail about AI’s role or non-role in incidents in the future.”
Our reading: The public may get little detail about AI's role in future military incidents.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“We should absolutely want any AI that is used for war to be as precise and reliable as possible, minimizing casualties, but we should also never forget that those who use them are responsible for decisions about how many casualties are acceptable.”
Our reading: We should absolutely want any AI used for war to be as precise and reliable as possible, minimizing casualties, and we should never forget that those who use it are responsible for decisions about how many casualties are acceptable.
No horizon
Scope: any AI that is used for war
Assessment: Can’t be checkedSaid
Observation
“Whether or not current AI algorithms are precise (probably they aren’t), and whether not humans are involved in the specific selection of targets, those who use military algorithms bear responsibility for the outcomes they produce.”
Our reading: Current AI algorithms are probably not precise.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Whether or not current AI algorithms are precise (probably they aren’t), and whether not humans are involved in the specific selection of targets, those who use military algorithms bear responsibility for the outcomes they produce.”
Our reading: Whether or not current AI algorithms are precise (probably they aren't), and whether or not humans are involved in the specific selection of targets, those who use military algorithms bear responsibility for the outcomes they produce.
No horizon
Qualifier: whether or not current AI algorithms are precise (probably they aren't)
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The race to shove AI into everything is grossly premature, because the tech fundamentally lack reliability.”
Our reading: The race to shove AI into everything is grossly premature because the technology fundamentally lacks reliability.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“Meanwhile, the chance that we will get straight answers is probably close to zero.”
Our reading: The public will very probably not get straight answers about AI's role in military incidents.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Anthropic deserves a chance at EXACTLY the same terms; anything else reeks of corruption.”
Our reading: Anthropic should be given a chance at exactly the same government terms as OpenAI; anything else would look corrupt.
No horizon
Source: The whole thing was a scam
Assessment: Can’t be checkedSaid
Rhetoric
“It was all theatre. Dario never had a chance.”
Our reading: The whole episode, including Altman's public support for Amodei while he was secretly working on his own deal, was all theatre, and Dario never had a chance.
No horizon
Source: The whole thing was a scam
Assessment: Can’t be checkedSaid
Forecast
“My guess is that in time Trump will be remembered more for his AI policies than for ICE or his tariffs, and not in a good way.”
Our reading: Gary Marcus guesses that, in time, Trump will be remembered more for his AI policies than for ICE or his tariffs, and negatively.
No horizon
Qualifier: more than for ICE or tariffs
Assessment: Can’t be checkedSaid
Observation
“This will almost certainly alienate a large part of Silicon Valley, and puts Sam Altman, and indeed the whole Valley, in a serious bind.”
Our reading: Trump's doubling down against Anthropic puts Sam Altman and the whole of Silicon Valley in a serious bind.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The finances just don’t make sense.”
Our reading: The finances behind OpenAI's new financing just don't make sense.
No horizon
Assessment: Can’t be checkedSaid
Forecast
“Even with Nvidia and other propping them up, I hold onto my longstanding sense that there is a good chance that OpenAI will someday be seen as the WeWork of AI.”
Our reading: There is a good chance that OpenAI will eventually be seen as the WeWork of AI, even with Nvidia and others supporting it.
No horizon
Qualifier: even with Nvidia and others propping it up
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“That OpenAI’s’s valuation should more or less double in a year or so when they all but squandered their former lead appears, at least to me, to be insane.”
Our reading: To Marcus at least, it appears insane that OpenAI's valuation should more or less double in a year or so when OpenAI has all but squandered its former lead.
No horizon
Qualifier: appears, at least to me
Assessment: Can’t be checkedSaid
Forecast
“This could well extend to nuclear weapons.”
Our reading: Unrestricted military use of Anthropic's AI could well extend to nuclear weapons.
No horizon
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Causal claim
“Which means we might not have a Stanislav Petrov next time around.”
Our reading: Because the Block Nuclear Launch by Autonomous AI Act failed, there might be no human like Stanislav Petrov to stop a mistaken nuclear launch next time.
No horizon
Qualifier: given the failure of the Block Nuclear Launch by Autonomous AI Act
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Mass surveillance and AI-fueled weapons, possibly nuclear, without humans in the loop are categorically not things that one individual, even one in the Cabinet, should be allowed to decide at a gunpoint.”
Our reading: Mass surveillance and AI-fueled weapons, possibly nuclear, without humans in the loop are categorically not things that one individual, even one in the Cabinet, should be allowed to decide at gunpoint.
No horizon
Qualifier: possibly nuclear
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“He is right.”
Our reading: Marcus endorses the reader's view that whether AI may be used for surveillance of Americans and lethal strikes without humans in the loop must be decided publicly by Congress, not forced by the Pentagon.
No horizon
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Observation
“Nothing that I have read convinces me that Secretary Hegseth has a nuanced understanding of the strengths and limits of current AI, or that he will show restraint in how he applies it.”
Our reading: Nothing Marcus has read convinces him that Secretary Hegseth has a nuanced understanding of the strengths and limits of current AI, or that Hegseth will show restraint in how he applies it.
No horizon
Scope: nothing that I have read
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Observation
“The second is that Hegseth’s maneuver is an audacious power grabs that aims to circumvent Congress. By setting a deadline of 5:01 PM eastern tomorrow, Hegseth aims to cut everybody else – even Congress — out of the loop.”
Our reading: Hegseth's deadline to Anthropic is a power grab meant to circumvent Congress and cut others out of the decision.
No horizon
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“AI policy, especially of this magnitude, is something that American people should have a say in. Congress should deliberate.”
Our reading: The American people should have a say in AI policy, especially AI policy of this magnitude, and Congress should deliberate.
No horizon
Qualifier: especially of this magnitude
Source: America, and probably the world, stands on a precipice.
Assessment: Can’t be checkedSaid
Conditional forecast
“Spreading it everywhere without proper consideration could well lead to disaster.”
Our reading: Deploying LLMs everywhere without proper consideration of their unreliability could well lead to disaster.
No horizon
Only if: if spread everywhere without proper consideration
Source: 30 seconds to midnight? 15?
Assessment: Can’t be checkedSaid
Conditional forecast
“It should be obvious that applying hallucinatory and unreliable Generative AI to weapon systems without humans in the loop could be catastrophic.”
Our reading: Applying generative AI to weapon systems without humans in the loop could be catastrophic.
No horizon
Only if: if applied to weapon systems without humans in the loop
Source: 30 seconds to midnight? 15?
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“If we’re going to embed LLMs into the fabric of the world - and apparently we are - we MUST do so in a way that acknowledges and factors in their unreliability.”
Our reading: If LLMs are going to be embedded throughout the world, this must be done in a way that recognizes and accounts for their unreliability.
No horizon
Qualifier: if LLMs are embedded throughout the world
Source: 30 seconds to midnight? 15?
Assessment: Can’t be checkedSaid
Rhetoric
“Given White House pressure to immediately use models that we cannot trust throughout the military, this is no longer a drill.”
Our reading: With the White House pushing to use untrustworthy AI models across the military right away, the danger is now real.
No horizon
Qualifier: given White House pressure to use untrusted models throughout the military
Source: 30 seconds to midnight? 15?
Assessment: Can’t be checkedSaid
Conditional forecast
“But unless Anthropic successfully stares down the Department of War, generative AI, as “jagged” as it is, will likely soon be used in *exactly* that way.”
Our reading: Unless Anthropic successfully resists the Department of War, generative AI will likely soon be used to make major life-or-death decisions.
No horizon
Only if: unless Anthropic successfully resists the Department of War
Source: 30 seconds to midnight? 15?
Assessment: Can’t be checkedSaid
Rhetoric
“We are on a collision course with catastrophe.”
Our reading: The current path of AI deployment, especially in the military, is heading toward catastrophe.
No horizon
Source: 30 seconds to midnight? 15?
Assessment: Can’t be checkedSaid
Conditional forecast
“And it may well lay waste to our economy, once banks and investors who bought the hype start to fall.”
Our reading: Generative AI may well lay waste to the economy once banks and investors who bought the hype start to fail.
No horizon
Only if: once banks and investors who bought the hype start to fall
Assessment: Can’t be checkedSaid
Forecast
“When all is said and done, my best guess is that generative AI will have done significantly more harm to society than good.”
Our reading: Marcus's best guess is that generative AI will ultimately have done significantly more harm to society than good.
No horizon
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Confusing increasingly sophisticated statistical approximation with general intelligence is therefore not only a conceptual error, but a strategic misjudgment.”
Our reading: Confusing increasingly sophisticated statistical approximation with general intelligence is not only a conceptual error but also a strategic misjudgment.
No horizon
Source: Rumors of AGI’s arrival have been greatly exaggerated
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“By focusing on specific benchmarks, these definitions miss the generality that was central to the original definitions, as discussed in the next section, thereby reflecting a fundamental misunderstanding of how intelligence should be evaluated.”
Our reading: Definitions that treat AGI as strong performance on standardized benchmarks miss the generality that was central to the original definitions, and so reflect a fundamental misunderstanding of how intelligence should be evaluated.
No horizon
Source: Rumors of AGI’s arrival have been greatly exaggerated
Assessment: Can’t be checkedSaid
Causal claim
“As AI systems become embedded in scientific and institutional decision-making (NIST, 2023; European Commission, 2024), overestimating their cognitive capacities risks misallocating trust, responsibility, and authority.”
Our reading: Overestimating AI's cognitive capacities as it is embedded in decision-making risks misallocating trust, responsibility, and authority.
No horizon
Qualifier: as AI becomes embedded in decision-making
Source: Rumors of AGI’s arrival have been greatly exaggerated
Assessment: Can’t be checkedSaid
Observation
“Here, we argue that recent claims of achieving AGI rest on a conceptual error: conflating increasingly sophisticated statistical approximations with intelligence itself.”
Our reading: Recent claims of achieving AGI conflate sophisticated statistical approximation with intelligence itself.
No horizon
Source: Rumors of AGI’s arrival have been greatly exaggerated
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“No use of the first person by chatbots, and no more deepfakes of living people’s voices and images without their express consent, aside from carveouts for obvious parody and so on.”
Our reading: Federal law must bar chatbots from using the first person and bar any further deepfakes of living people's voices and images without their express consent, with exceptions for obvious parody and other similar cases.
No horizon
Qualifier: without their express consent
Source: We URGENTLY need a federal law forbidding AI from impersonating humans
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“The need for the kind of law he was calling for – one forbidding “the creation of and ‘passing along’ of counterfeit people” is now urgent.”
Our reading: A law forbidding the creation and passing along of AI 'counterfeit people' is now urgently needed.
No horizon
Source: We URGENTLY need a federal law forbidding AI from impersonating humans
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“And we must not let corporate lobbyists thwart efforts to address all this.”
Our reading: Corporate lobbyists must not be allowed to thwart efforts to regulate AI impersonation and deepfakes.
No horizon
Source: We URGENTLY need a federal law forbidding AI from impersonating humans
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Tell your representatives today, not tomorrow, that we must pass federal1 laws forbidding machine output from being presented as humans, and that we must develop the means to enforce those laws.”
Our reading: The US must pass federal laws forbidding machine output from being presented as human and develop the means to enforce them.
No horizon
Source: We URGENTLY need a federal law forbidding AI from impersonating humans
Assessment: Can’t be checkedSaid
Recommendation or value judgment
“Generative AI systems may still struggle to reason, as discussed here recently, but they were built for mimicry, and their mimicry has gotten to the point where we must do something now.”
Our reading: Although generative AI systems may still struggle to reason, they were built for mimicry, and their mimicry has reached the point where we must do something now.
No horizon
Qualifier: may still struggle to reason
Source: We URGENTLY need a federal law forbidding AI from impersonating humans
Assessment: Can’t be checkedSaid
Forecast
“Small wonder OpenAI started probing about bailouts last fall; it’s looking more and more like they will need one.”
Our reading: OpenAI will likely need a government bailout.
No horizon