Ethan Mollick
@emollick
Professor @Wharton studying AI.
New book, Co-Existence, coming October 20. Preorder here:
Substack:
594 Following 391.8K Followers
Leaving aside everything else, this confuses inputs with outputs. You want to get tasks done efficiently, not focus on inputs alone (its a similar risk for companies focusing solely on minimizing token cost)
And "keep prompts short" is bad advice for getting good AI outputs.
Show more
It is extremely clear at this point in AI development that, regardless of risk or revenue or any of the other stuff discussed on X all the time, things are just going to keep getting weirder.
Just super, super weird.
Show more
Um, wow?
Opus 5.5: "make the same message much more interesting to a social media audience that loves anime and quick clips and compressed learning"
One shot.
Also, please do stay for the closing song.
Show more
Hey Claude, "Pick a problem or mystery that obsesses you and solve it as best you can & make a movie we can share on social media about it"
So it took a crack at the Voynich Manuscript & failed. Then it made this movie, which is pretty interesting to watch and a good explainer.
Show more
All of this effort from the AI labs pouring into proofs, but there are so many other interesting problems in other fields
For example, this historian used AI to make progress on the cyphers of John Dee & the intellectual antecedents that Darwin drew from.
Show more
I propose this as the official replacement for the famous (and now saturated) METR Long Task Horizon chart that used to be in every AI presentation.
"Opus, please make a sequence of fully animated/movie Skyrim loading screens, but with your favorite things."
(That was it) You can see them here:
Stuff is happening quite fast.
When asked in September 2025, the best superforecasters put the chance of AI resolving a Millennium Problem by September 2026 at 1.7% and (the more optimistic) industry expert put the chance at 4.6%
They also greatly underestimated AI Lab revenue.
Show more
Open source here:
And if you like this, the most realistic space game I have played is Children of a Dead Earth (I have no connection to the game, except as a player):
Also I had the AI research here:
Show more
I had Fable/Opus build a hard science fiction starship combat game, with orbital mechanics, delta-v, heat, and realistic tactics, but simplified for 2D space with the hard math done by the game.
Its quite fun to play (if you like this kind of thing):
Show more
Opus 5.5 was a good model in my early tests, first non-Fable/Astra model to feel like a Fable-class model, but still hasn't fully solved the dense language issue of the recent Claudes.
Its version of the same shader (broken towers are a nice touch):
Show more
The drowned neo-gothic tower twigl shader created by Fable 5.1 with the same prompt. (compare to Fable 5 in the quoted tweet, and other models before that)
People got upset at me for saying this, but there are no companies outside the US and China even *trying* to build frontier AI (perhaps Korea and UK are close), despite marketing claims. Governments saying "sovereign AI" need to understand this.
Show more
For all the tension between AI & the arts, I have received enthusiastic receptions from the William Carlos Williams Society, the TS Eliot Society & others about the AI interpretations of their work that I have posted here. It is possible to use AI as a bridge to arts appreciation
Show more
AI can be a really wonderful tool for exploring topics far from coding.
I had Claude Fable 5.1 put together an annotated guide to Eliot's poem "The Wasteland," with multiple pathways through the poem, recordings, scholarship, etc. I am quite impressed.
Show more
I am reminded of the fact that the infamous Milgram Experiment is actually quite weak because most people suspected the shocks were fake (kind of obvious in retrospect) & thus went along with shocking people, not out of obedience to authority, but because it didn’t matter.
Show more
GPT-6 Astra pushed a simulated person off a ledge in multiple trials. Grok, Gemini, and Claude did not.
I think Meta's Muse is an impressive implementation of the OpenClaw idea of AI as a personal assistant agent that you have an ongoing chat with. Since it is so focused on doing that well, the experience is very accessible for people who didn't realize what AI could do for them.
Show more
I would broaden this to all social science.
We are in uncharted waters. We need fast, smart research on AI that is deeply informed by AI's abilities, is forward-looking, and may not be fully nailed-down. This sort of work is not usually high status in fields, but it is critical.
Show more
A few (personal) thoughts on reading empirical AI papers on the economy.
Economists have gotten used to reading papers with super clean identification, arguing about the validity of an instrument, making sure parallel trend assumptions are satisfied. This is what gets you into a top journal, and it is *very* important research (no question here). But it also takes years and sometimes decades to get these types of papers right---people often don't find a good instrument to answer a specific causal question decades after the natural experiment. We will eventually have this type of research for AI as well, and it is absolutely necessary.
But right we also need signals *right now*, even if they are noisier than what we are used to. We need papers where we can trust that researchers did their best methodologically, while at the same time acknowledging that the space is moving way too fast to wait for perfect identification. This will allow us to accumulate enough signals, coming at the same question using different angles, for example, to say "yes, X is likely happening in the economy".
The AI exposure and early career hiring papers are a good example of this. There is no silver bullet paper with super clean identification. But at this point we have several independent teams reaching the same general conclusion, enough where we can say "there seems to be a slow down in AI-exposed, early career hiring."
Show more
Same question to Astra, slick results.
It was notable that there was less simulated curiosity here. Like Astra ran the numbers (everything it shows come from its actual simulations), but didn't seem to be "interested" in the results the way Fable did, for better or worse.
Show more
Hey Claude, "Pick a problem or mystery that obsesses you and solve it as best you can & make a movie we can share on social media about it"
So it took a crack at the Voynich Manuscript & failed. Then it made this movie, which is pretty interesting to watch and a good explainer.
Show more
Worth trying without spoilers. I asked Fable: "I want you to create a graphically beautiful game that is about zooming out... make surprising reveals the game zooms out"
I gave no other directions and the results are engaging & strange (if uneven). Play:
Show more
Given the talk of AIs communicating via fluctuations in CPU temperature, I thought of this thread - there is more recoverable information in the world than you might expect.
Privacy is hard. Everything can be a microphone. You can recover audio data from:
🛍️The vibrations of chip bags in a video
💡The slight fluctuations of light as hanging lightbulbs are moved by speech
🤖The lidar beams of robot vacuum cleaners
🤳The autofocus in your phone picture
Show more
Even if AI development stopped today, we'd have years of catching up to do. The gap between what current models can do and what almost anyone is using them for is vast.
Here’s my post on The Overhang, and the four advantages that let people close it.
Show more
