25 AI Predictions for 2025, from Marcus on AI

With a review of last year’s predictions | Continue reading


@garymarcus.substack.com | 16 hours ago

Where will AI be at the end of 2027? A bet

We, Gary Marcus, author, scientist, and noted AI skeptic, and Miles Brundage, an independent AI policy researcher who recently left OpenAI and is bullish on AI progress, have agreed to the following bet, at 10:1 odds, with criteria drawn from two earlier Substack essays by Gary t … | Continue reading


@garymarcus.substack.com | 2 days ago

𝗼𝟯 “𝗔𝗥𝗖 𝗔𝗚𝗜” 𝗽𝗼𝘀𝘁𝗺𝗼𝗿𝘁𝗲𝗺 𝗺𝗲𝗴𝗮𝘁𝗵𝗿𝗲𝗮𝗱: 𝘄𝗵𝘆 𝘁𝗵𝗶𝗻𝗴𝘀 𝗴𝗼𝘁 𝗵𝗲𝗮𝘁𝗲𝗱, 𝘄𝗵𝗮𝘁 𝘄𝗲𝗻𝘁 𝘄𝗿𝗼𝗻𝗴, 𝗮𝗻𝗱 𝘄𝗵𝗮𝘁 𝗶𝘁 𝗮𝗹𝗹 𝗺𝗲𝗮𝗻𝘀

Kevin Roose, of Hard Fork and NYT, was so impressed with OpenAI’s rollout that he joked “of course they have to announce AGI the day my vacation starts”. | Continue reading


@garymarcus.substack.com | 10 days ago

o3, AGI, the art of the demo, and what you can expect in 2025

OpenAI’s new model was revealed yesterday; its most fervent believers think AGI has already arrived. Here’s what you should pay attention to in the coming year. | Continue reading


@garymarcus.substack.com | 11 days ago

Generative AI’s Continuing Copyright Problems, an Essay in Memory of Suchir Balaji, 1998 - 2024

In early November, I had a stimulating Zoom call with a former OpenAI employee and Berkeley graduate named Suchir Balaji, who had just left OpenAI. | Continue reading


@garymarcus.substack.com | 18 days ago

Humanity’s “Oh shit!” AI moment?

Not yet, but it could come sooner than you think. Not because we are close to AGI, but because we already have machines that can say one thing and do something else altogether. | Continue reading


@garymarcus.substack.com | 20 days ago

Casey (Newton) Strikes Out

His continued skepticism of AI skepticism continues to miss the mark | Continue reading


@garymarcus.substack.com | 21 days ago

Sora still appears to have trouble with physics

Exactly as I warned in February | Continue reading


@garymarcus.substack.com | 23 days ago

Don’t Ride This Bike! Generative AI’s persistent trouble with compositionality and parts

When the text-to-image AI generation system DALL-E2 was released in April 2022, the two of us, together with Scott Aaronson, ran some informal experiments to probe its abilities. | Continue reading


@garymarcus.substack.com | 24 days ago

Hard-forked! Casey Newton’s distorted portrait of Gary Marcus and AI skepticism

What Casey Newton’s new essay skewering AI skepticism does—and does not—get right | Continue reading


@garymarcus.substack.com | 26 days ago

Two visions of AI’s future

It’s not looking good | Continue reading


@garymarcus.substack.com | 27 days ago

Which CEO will be the last to see it?

What kept me going during the dark years of 2022 and 2023 when Generative AI was wildly overhyped and I was constantly ridiculed was the secure knowledge that the truth would eventually come out. | Continue reading


@garymarcus.substack.com | 27 days ago

ChatGPT, at Age Two

The bullshit just keeps on coming. | Continue reading


@garymarcus.substack.com | 1 month ago

Cognitive scientist Gary Marcus says AI must be regulated. He has a plan.

Fantastic writeup of my views today at The Wall Street Journal: | Continue reading


@garymarcus.substack.com | 1 month ago

Update re: Microsoft and training data

Relevant to the post I sent earlier today, https://www.howtogeek.com/is-microsoft-using-your-word-documents-to-train-ai/says that an unnamed spokesperson at Microsoft claims that “Microsoft does not use customer data from Microsoft 365 consumer and commercial applications to trai … | Continue reading


@garymarcus.substack.com | 1 month ago

PSA: Microsoft may be training on your private data without your knowledge

The Great Data Heist continues. | Continue reading


@garymarcus.substack.com | 1 month ago

A new AI scaling law shell game?

Scaling laws were supposed to be, well, laws. | Continue reading


@garymarcus.substack.com | 1 month ago

Satya Nadella and the three stages of scientific truth

A textbook case in how good ideas are often initially throttled | Continue reading


@garymarcus.substack.com | 1 month ago

On hype, and the unbearable banality of ChatGPT’s poetry

A new AI study is making the rounds, claiming that ChatGPT can write poetry that is “indistinguishable” from William Shakespeare. | Continue reading


@garymarcus.substack.com | 1 month ago

An AI rumor you won’t want to miss

Converging evidence that the core hypothesis driving generative AI may be wrong | Continue reading


@garymarcus.substack.com | 1 month ago

CONFIRMED: LLMs have indeed reached a point of diminishing returns

Science, sociology, and the likely financial collapse of the Generative AI bubble | Continue reading


@garymarcus.substack.com | 1 month ago

8 signs that Donald Trump has a progressive form of dementia

And how our media is failing, wildly, in one of the most important stories in US history | Continue reading


@garymarcus.substack.com | 2 months ago

AI influencer hype gone wild

Reality is quite the opposite | Continue reading


@garymarcus.substack.com | 2 months ago

An epidemic of weirdly precise AI butt facts

The parade of made-up numbers never stops | Continue reading


@garymarcus.substack.com | 2 months ago

Is OpenAI more like WeWork or Theranos?

A tale of two analogies | Continue reading


@garymarcus.substack.com | 2 months ago

When it comes to security, LLMs are like Swiss cheese — and that’s going to cause huge problems

Two new examples that are quite disconcerting | Continue reading


@garymarcus.substack.com | 2 months ago

LLMs don’t do formal reasoning - and that is a HUGE problem

Important new study from Apple | Continue reading


@garymarcus.substack.com | 2 months ago

Two Nobel Prizes for AI, and Two Paths Forward

As you have surely read by now, not one but two Nobel Prizes went to AI this week. | Continue reading


@garymarcus.substack.com | 2 months ago

Eric Schmidt’s Risky Bet on AI and Climate Change

Is AI so valuable to society that we should place this bet? | Continue reading


@garymarcus.substack.com | 2 months ago

Sorry, GenAI is NOT going to 10x computer programming

Here’s Why | Continue reading


@garymarcus.substack.com | 3 months ago

How much should OpenAI’s abandoned promises be worth?

Once upon a time OpenAI promised to be a non-profit, pledged to public benefit. | Continue reading


@garymarcus.substack.com | 3 months ago

Five reasons why OpenAI’s $150B financing might fall apart

And what that might mean for Generative AI more broadly | Continue reading


@garymarcus.substack.com | 3 months ago

OpenAI’s slow-motion train wreck

Mira Murati’s departure is the just the last episode in a long-running drama | Continue reading


@garymarcus.substack.com | 3 months ago

Gary Marcus vs Sam Altman on GPT-o1 aka Strawberry

You might be surprised | Continue reading


@garymarcus.substack.com | 3 months ago

What will the candidates say about AI in tonight’s debate?

AI policy *should* be front and center | Continue reading


@garymarcus.substack.com | 3 months ago

Scoop: What former employees of OpenAI are worried about

An inside look. And, spoiler alert, it’s not Q* | Continue reading


@garymarcus.substack.com | 4 months ago

Why California’s AI safety bill should (still) be signed into law - and why that won’t be nearly enough

Thursday broke my heart. | Continue reading


@garymarcus.substack.com | 4 months ago

What has and has not changed in the AI since the ChatGPT revolution? [Video]

Video of a new talk | Continue reading


@garymarcus.substack.com | 4 months ago

An open letter to Fei-Fei Li concerning California’s proposed AI regulation, SB-1047

Dear DrFeiFei, | Continue reading


@garymarcus.substack.com | 4 months ago

OpenAI: On a Path to Becoming The World’s Most Frightening Surveillance Company?

⚠️ Something terrifying just snapped in place: | Continue reading


@garymarcus.substack.com | 4 months ago

The OpenAI Plot Thickens

Crazy day gets crazier | Continue reading


@garymarcus.substack.com | 4 months ago

August 5, 2024, a big day in tech?

Big day. | Continue reading


@garymarcus.substack.com | 4 months ago

Why the collapse of the Generative AI bubble may be imminent

An update from the person who first called the bubble | Continue reading


@garymarcus.substack.com | 5 months ago

This one important fact about current AI explains almost everything

An entire industry has been built - and will collapse - because many people haven’t yet gotten it. | Continue reading


@garymarcus.substack.com | 5 months ago

Five signs that the GenAI honeymoon is over

And why that might be a good thing | Continue reading


@garymarcus.substack.com | 5 months ago

Most people don’t think GenAI has improved their productivity

Damning new study from Upwork | Continue reading


@garymarcus.substack.com | 5 months ago

AlphaProof, AlphaGeometry, ChatGPT, and why the future of AI is neurosymbolic

What comes after chatbots? | Continue reading


@garymarcus.substack.com | 5 months ago

Why OpenAI may well be completely Zuck’d

Regular readers of this Substack will recall that I have never been bullish on OpenAI. Last Fall, I warned that they might someday be known as the WeWork of OpenAI, and in late January I wrote an essay about some of the strong headwinds they faced, despite their seeming invincibi … | Continue reading


@garymarcus.substack.com | 5 months ago