The TLDR seems perhaps tongue-in-cheek, but it's a worthy thought: the situation where our best experts find new science to be inscrutable is close at hand, if not here now.
Humans have been guilty of burying things we don't understand for millenia. Will we continue to do so "because AI"?
I really love this idea given the recent controversy around mathematics solutions.
It seems like a "mere suggestion" of success has a positive impact on finding solutions.
We know this technique works in humans, from which this is all derived from, so it seems to make sense.
Thank you. This is what I'm driving at, that most of the AI and software devs here seem to be missing. Intelligence is not, and never was the bottleneck for most science/hard tech. Full AGI gets, at best, a small productivity improvement, which over long periods of time does have compounding effects. But this isn't a singularity hard-takeoff inflection point.
How much lab equipment is automatable though? There's definitely some in biology, but if you're doing fundamental research it's 99% stuff you are building yourself with your own hands. Robotics is a long way from being able to do any of that.
When I was doing research (physical electronics, lasers, fiberoptics and sensors stuff), lot of time was spent just writing all sorts of DAQ and processing code. So all this LLM stuff would have been really useful. There's a lot of data collection, data processing in the lab that require all sorts of ad hoc scripts and stuff.
That was many years ago, but I would be surprised if the current crop of researchers are not using these things. And if they are not, then they are just not serious.
>but if you're doing fundamental research it's 99% stuff you are building yourself with your own hands.
You can do LLM->3D Printed models now. The drone can fly in and pick them up and bring them to the location you want. They can assemble structures. All automated, all LLM driven.
that might indeed be a problem for all the pulp-producing labrats of STEM in southern europe and the third world.
However I think this area has so much decoupled from industry and solid research institutions that they might not notice at all (beyond their use of AI-generated slop to augment the slop they already produce)...
1. I can't effectively engage in follow-up conversation with the "author." I can't do it with the human (principal) because they didn't write it and can't explain it--and, in fact, they may not even personally agree with all if it!
2. It shifts the cognitive burden from writer to reader. What makes writing hard, and valuable, is the effort put into translating one's raw, unfiltered thoughts into writing that is easy to comprehend, and the writer's own unique personality that is expressed in it. As ambient prose, AI-generated writing is often not only more difficult to read than human-drafted prose, but it also has this "sameness" quality that makes every "author" have the same personality.
The purpose of writing is to communicate. The person communicating is an important part of the message.
Consider the sentence "I'm waiting for you at home." It really matters if it's said by your loving wife or an evil clown. You can extrapolate from this.
Imagine a future where 99% of text is written by 1 unpassionate person. Reading the same patterns, the same tropes, the same flow across paragraphs, over and over, would be unpleasant. It already is unpleasant.
The problem is that some people think all AI-assisted writing is prompting "write me a post about X", so they see LLM prose and assume there has been no thought or effort put into it.
So how much time do you deem a minimum amount for someone to spend on some unit of work before you're willing to read it? And what technologies are allowed -- are word processors allowed or does it have to be done with a writing instrument like a pencil? What about spell check? Do you care if they had done their research with a Wikipedia or do you prefer card catalog for the really important work?
I'm not sure. Another way to think about it: how _little_ time is someone able to spend on something where you will still be interested in reading it? If someone tells their unattended LLM agent to create a blog, write whatever content will maximize banner ad revenue, and then share each post on Hacker News, would you read it? What if the content is actually interesting in your opinion?
For me, the amount of time could be very minimal. A person can splash some paint on a canvas, tape a banana to a wall, press a shutter, or write 14 syllables and if it's well considered then that can be enough. Time spent just doesn't play into it for me, unless that's actually part of the point.
> If someone tells their unattended LLM agent to create a blog, write whatever content will maximize banner ad revenue, and then share each post on Hacker News, would you read it? What if the content is actually interesting in your opinion?
Yes, I think if one of that content had a headline that interested me I would definitely click on it, because that's already happened!
In one case, I could tell because the tone of the piece sounded exactly like the kind of stuff I would read after prompting an LLM and I had never heard of the author before, but I knew that they were like an intern or a new college grad and for some reason their piece was trying to persuade employees in a certain field about how they had to use AI itself. It just rubbed me the wrong way because if they did use AI, then they didn't even bother to rewrite the repetitive instances of "it's not X; it's y", and also "This isn't X".
Because what the clanker produces is almost always bad output compared to what a capable human would produce. Furthermore, even an incapable human is better than a clanker, because at least humans have intrinsic value that they bring to the table. Finally, it's just plain disrespectful to refuse to take some time to answer a question and hand it off to the clanker.
A single counter example: In East Bay CA. I designed and built my 9kw solar roof installation myself. No electricians at all. The process was very easy and working with my city and power company was simple, with no blockers at all. 4 years running. On a yearly basis, we make over double what we use. It boggles my mind how solar isn't on every house in the country.
Humans have been guilty of burying things we don't understand for millenia. Will we continue to do so "because AI"?
reply