Don't Trust the Science

There is no actual science about AI yet

There are a lot of absurd pseudo-scientific studies floating around now that scientists have discovered that putting “AI” in the title significantly increases the chances that they’ll get a government grant. Don’t pay any attention to them whatsoever, and certainly don’t doubt the evidence of your own experience with AI.

Who cares if AI supposedly makes one group of workers more efficient while reducing another group of students to drooling psychopaths fantasizing about murdering their teacher. It’s all nonsense, and all that matters is how your utilization of AI enhances or degrades your own personal workflow.

I wrote and recorded about 30 songs without AI in 30 years. I wrote and recorded about 70 songs with AI in nine months. Suffice it to say that I’m a big fan of AI music in general and Suno in particular.

But your mileage may vary. Consider this very well-regarded study published by a student at MIT.

Toner-Rodgers: So my first result is just looking, on average, at how this tool impacted both the discovery of new materials as well as downstream innovation in terms of patent filings and product prototypes. So I find that researchers with access to the AI tool discover 44 percent more materials, and then this results in a 39 percent increase in patent filings and then a 17 percent rise in downstream product innovation, which I measure using the creation of new product prototypes that incorporate those materials.

Demsas: These are, like, massive numbers.

Toner-Rodgers: Yeah, I think they’re pretty big. And also, I think it’s helpful to kind of step back and look at the underlying rate of productivity growth in terms of the output of these researchers. So I look back at the last five years before the tool was introduced, and output per researcher had actually declined over this period. So these are huge numbers relative to the baseline rate of improvement.

Demsas: So it’s interesting—well, I guess first: How? Like, why are people becoming more productive here?

Toner-Rodgers: I think there’s two things. So one is just that the tool is pretty good at coming up with new compounds. So being able to train a model on a huge set of existing compounds is able to give a lot of good suggestions.

And then second: Not having to do that compound design part of the process themselves frees scientists to spend more time on those second two categories, kind of deciding which materials to test and then actually going and testing their properties.

Demsas: It’s interesting when I was looking at your results because you’re able to kind of look at, you know, one month after, four months after the adoption of this new AI tool, how it changes things. Things look kind of grim in the short run, right? Like, four months after AI adoption, the number of new materials actually drops. And it’s not until eight months after that you see a significant increase in new materials. And that’s around when you see the patent filings increase. And it’s not until 20 months after that you actually see it show up in product prototypes.

And, you know, part of the problem of trying to figure out if new technology like AI is having a big impact is that it might take a while to show up in statistics. Is that why you think maybe we’re not seeing a massive jump in productivity right now in the U.S., despite the rollout of a ton of new machine-learning tools?

Toner-Rodgers: Yeah, I think that’s partly true. Like, you definitely need some forms of organizational adaptation or people learning to actually utilize these tools well. So part of why there’s this lag in the results is just that materials discovery takes a while. So it takes a little bit to actually go and kind of synthesize these compounds and then go and find their properties.

But another thing I find is that in the first couple months after the tool’s introduction, scientists are very bad, across the board, at determining which of the AI suggestions are good and which are bad. And this is part of the reason we don’t see effects right away.

Demsas: So it’s like your job has changed significantly, and you just need time to adjust to that.

Toner-Rodgers: Yeah, totally.

Powerful stuff, right? 39 percent more patent filings and 17 percent more product innovation! Wow, this new technology is going to have a massive positive impact on… whatever the particular application is, right?

Not so fast.

Aidan’s paper cited six articles by Acemoglu and another six by Autor, which likely contributed to why they were so impressed.

Despite only being a preprint, it has already been cited over 60 times. The paper was also featured in Nature, NPR, Freakonomics, Marginal Revolution, cited by the European Central Bank, and even in the U.S. Congress.

Here’s Toner-Rodgers presenting the paper at a research seminar:

• https://cassyni.com/events/MiPYGu3qzKP5MQFWNUn9Tb

Aidan then did a podcast circuit to promote the paper, including a full hour with The Atlantic:

• https://www.theatlantic.com/podcasts/archive/2025/01/ai-scientific-productivity/681298/

Just a few months after his media tour, Aidan’s paper was revealed as a fraud: not just fraudulent, but *spectacularly* fraudulent. The entire study was fabricated from thin air, and yet somehow, this psycho gave eloquent seminars and spoke about it for an entire hour to the Atlantic.

The bombshell exposing him as a fraud was dropped 2 months ago when the WSJ — the very same newspaper that honored him with a glowing photoshoot just 4 months prior — published a story: ‘‘MIT Says It No Longer Stands Behind Student’s AI Research Paper’’

Due to the corruption of science, it’s effectively worse than useless these days; a literal coin flip is more reliable than the average published, peer-reviewed scientific study. So don’t pay any attention to whatever it is the grant-chasers are saying now, just get your hands dirty, experiment, and find out for yourself whether AI works for you or not.