An artist fights back, and Midjourney has embarrassed themselves

Reid Southen is a successful concept artist who has worked for many of the biggest studios (Marvel, 20th Century Fox, Warner Bros, Paramount etc) on a lot of huge films (Matrix Resurrections, The Hunger Games, Transformers, and Alien, among others). | Continue reading


@garymarcus.substack.com | 1 year ago

Are social media and coordinated misinformation campaigns fomenting extremism?

Probably, yes. | Continue reading


@garymarcus.substack.com | 1 year ago

Sorry, but FunSearch probably isn’t a milestone in scientific discovery

As ever, don’t believe the hype | Continue reading


@garymarcus.substack.com | 1 year ago

Meta’s Chief AI Officer is lying about the EU AI Act

A few days ago it looked like the EU AI Act had finally been negotiated, and resolved. There was much rejoicing. But the big tech executives that would need to deal with it aren’t happy, and are doing their best after the fact to undermine a hard-won compromise that involve … | Continue reading


@garymarcus.substack.com | 1 year ago

Nope, the Turing Test has not been solved

though it might not matter that much if it did | Continue reading


@garymarcus.substack.com | 1 year ago

It was always about trust

Big scoop from the NYT [Tripp Mickle, Cade Metz, Mike Isaac and Karen Weise] Maybe this is the key example? | Continue reading


@garymarcus.substack.com | 1 year ago

We interrupt this programming

Oh no. Oh no. Please tell me this isn’t true. And, ugh, it’s going to be impossible not to watch: § Business Insider’s scoop is here: https://www.businessinsider.com/openai-cofounder-ilya-sutskever-invisible-future-uncertain-2023-12 and here’s a juicy … | Continue reading


@garymarcus.substack.com | 1 year ago

The truth about Google’s amazing new demo

Friends don’t let friends take demos seriously | Continue reading


@garymarcus.substack.com | 1 year ago

Hot take on Google’s Gemini and GPT-4

Gemini was just announced. Here’s my hot take: Gemini seems to have by many measures matched (or slightly exceeded) GPT-4, but not to have blown it away. From a commercial standpoint GPT-4 is no longer unique. That’s a huge problem for OpenAI, especially post drama, w … | Continue reading


@garymarcus.substack.com | 1 year ago

“Not consistently candid”

Maybe OpenAI’s board had a point, after all | Continue reading


@garymarcus.substack.com | 1 year ago

Give me compromise, or give me chaos

The EU is about to make a decision that could have enormous, lasting repercussions. | Continue reading


@garymarcus.substack.com | 1 year ago

Critical national security questions should not be decided unilaterally by unelected tech leaders

Increasingly, they are. | Continue reading


@garymarcus.substack.com | 1 year ago

Hinton vs LeCun vs Ng vs Tegmark vs O

Three top ML researchers, a leading physicist, and a former French Minister, at each other’s throats | Continue reading


@garymarcus.substack.com | 1 year ago

AI versus The Clumsy Human Mind

Guess which one is better | Continue reading


@garymarcus.substack.com | 1 year ago

About that OpenAI “breakthrough”

Will OpenAI’s Q* change the world? | Continue reading


@garymarcus.substack.com | 1 year ago

A Tale of Two Liars

Actually, let me say at the outset, that my title is a literary stretch. Although this is a Dickensian tale of two that don’t tell the truth, one lies and the other doesn’t. The first is George Santos. He lies are legendary. His lies just got him convicted. | Continue reading


@garymarcus.substack.com | 1 year ago

Massive scoop

Massive Scoop from The Information: 👉Sam and Greg gave up their board seats 👉Sam agreed to an internal investigation This is inconsistent with all narratives about “cloddish” or self-interested boards, and suggests that something real happened. | Continue reading


@garymarcus.substack.com | 1 year ago

OpenAI is back in business

A few quick thoughts about what is really important | Continue reading


@garymarcus.substack.com | 1 year ago

Top 5 reasons why OpenAI was probably never really worth $86 billion

The real value was always the people, not the IP, not the data, not the customer list, not the infrastructure. Elon Musk’s Grok replicated much (not all) of what OpenAI had done in a few months; Kai-Fu Lee’s new company similarly was able to replicate a lot in less th … | Continue reading


@garymarcus.substack.com | 1 year ago

Open letter to all EU leaders

20 November 2023 Dear European leaders, The recent events at OpenAI are likely going to lead to considerable, unpredictable instability. The schisms on display there highlight the fact that we cannot rely purely on the companies to self-regulate AI, wherein even their own | Continue reading


@garymarcus.substack.com | 1 year ago

Everybody’s out

[resending to some readers, sorry for glitch] Posted this a few minutes ago, giving my theory of what transpired over the last few hours: and then, perhaps in keeping with my surmise that Sam’s bluff was called, negotiations broke down altogether. Sam’s out, presumabl … | Continue reading


@garymarcus.substack.com | 1 year ago

I’m not making this up

As Dave Barry used to say, I’m not making this up. Just passing it along. | Continue reading


@garymarcus.substack.com | 1 year ago

“No one person should be trusted here.”

The irony of OpenAI’s unusual structure | Continue reading


@garymarcus.substack.com | 1 year ago

As the OpenAI World Turns: 4 vital questions

Per The Verge, investors have asked the board to resign and want Sam Altman back in. They may well get their way. 4 vital questions: 👉Will the nonprofit continue to exist? 👉What checks and balances will there be? 👉Where will this leave OpenAI with respect … | Continue reading


@garymarcus.substack.com | 1 year ago

Firing Sam and then trying to bring him back is a hot mess.

Nobody told me there'd be days like these Nobody told me there'd be days like these Nobody told me there'd be days like these Strange days indeed Strange days indeed – John Lennon Sorry, me again. An hour after my last post, a credible rumor came out that OpenAI’s boa … | Continue reading


@garymarcus.substack.com | 1 year ago

What went down at OpenAI: A theory of the case

One theory of the case, popular on the social media platform X, is that the board was a bunch of clowns; they fired Sam for no really good reason beyond some ongoing tensions around company direction. Co-Founder Ilya Sutskever perhaps felt a bit left out. | Continue reading


@garymarcus.substack.com | 1 year ago

The biggest immediate lesson from the OpenAI meltdown

Altman is out. Hours later Greg Brockman, Co-Founder and Chairman of the Board resigned. More exits are rumored. Major drama at OpenAI. We don’t know what happened; many theories are swirling around. How things play out might or might have major impact on the prospects of g … | Continue reading


@garymarcus.substack.com | 1 year ago

Sam Altman appears to have been pushed out

Wow, just wow: "Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities” Marcus on AI is a reader-suppor … | Continue reading


@garymarcus.substack.com | 1 year ago

Has Sam Altman gone full Gary Marcus?

What a difference a year makes | Continue reading


@garymarcus.substack.com | 1 year ago

Crisis point in AI?

Corporate capture threatens to derail EU AI Act | Continue reading


@garymarcus.substack.com | 1 year ago

Rethinking “driverless cars”

Essentially every conversation about “driverless cars” over the last decade has to be rethought — with important implications as well for “AGI timelines”. | Continue reading


@garymarcus.substack.com | 1 year ago

Could Cruise be the Theranos of AI? And is there a dark secret at the core of the entire driverless car industry?

We don’t know, but here are some questions regulators should ask | Continue reading


@garymarcus.substack.com | 1 year ago

Where we are right now on open source and potential AI risk.

It’s not pretty | Continue reading


@garymarcus.substack.com | 1 year ago

The UK AI Summit

Glad to see something come out of it, but we could have hoped for more. | Continue reading


@garymarcus.substack.com | 1 year ago

First take on the White House Executive Order on AI

A great start that perhaps doesn’t go far enough | Continue reading


@garymarcus.substack.com | 1 year ago

Why driverless cars might—or might not—be at the end of the road

I lost my license now I don’t drive …. Life’s been good to me so far — Joe Walsh Breaking news: § If there’s been one person who has been even more skeptical about driverless cars than I have, going back almost a decade, it is the software engin … | Continue reading


@garymarcus.substack.com | 1 year ago

A CERN for AI and the Global Governance of AI

Forward motion is a good thing | Continue reading


@garymarcus.substack.com | 1 year ago

Getting GPT to work with external tools is harder than you think

A commonly-suggested fix isn’t working | Continue reading


@garymarcus.substack.com | 1 year ago

Hello, Multimodal Hallucinations

Broadening the training set doesn’t make Generative AI’s Achilles’ Heel go away | Continue reading


@garymarcus.substack.com | 1 year ago

How long does AI hype take to die?

A photoessay | Continue reading


@garymarcus.substack.com | 1 year ago

“Math is hard” — if you are an LLM – and why that matters

No matter how much data you train them on, they still don’t truly understand multiplication. | Continue reading


@garymarcus.substack.com | 1 year ago

Reports of the birth of AGI are greatly exaggerated

Breathless predictions about Artificial General Intelligence (AGI) are nothing new; they still aren’t true | Continue reading


@garymarcus.substack.com | 1 year ago

What if GPT-5 didn’t meet expectations?

It is quite possible that no prereleased product in the history of technology has ever has had more expectations put upon it | Continue reading


@garymarcus.substack.com | 1 year ago

Two kinds of Techno-Optimism

A few things Marc Andreessen forgot to mention | Continue reading


@garymarcus.substack.com | 1 year ago

Race, statistics, and the persistent cognitive limitations of DALL-E

Why DALL-E 3 is and isn’t better than DALL-E 2 | Continue reading


@garymarcus.substack.com | 1 year ago

What was 60 Minutes thinking, in that interview with Geoff Hinton?

Scott Pelley didn’t exactly do his homework | Continue reading


@garymarcus.substack.com | 1 year ago

Muddles about Models

Why it’s stretch to say that large language models represent ”literal world models” | Continue reading


@garymarcus.substack.com | 1 year ago

Seven lies in four sentences

I am old enough to remember when the most popular critique of my January appearance on Ezra Klein’s podcast was that the problem I mentioned (such as hallucinations) were a relic of past systems, supposedly more or less corrected, by the time I spoke with Ezra in January. S … | Continue reading


@garymarcus.substack.com | 1 year ago