4 ms·
My experience going through physics in undergrad + 1 year of grad school was that we had made enough progress that it takes ~8 years of book learning to get to
by lumost 4y ago
My experience going through physics in undergrad + 1 year of grad school was that we had made enough progress that it takes ~8 years of book learning to get to a basic competency at the forefront of a narrow part of physics. If you want to make a fundamental connection with a different field.. then adding enough knowledge to be able to formulate your hypothesis will take ~1-2 years of further book reading. Unfortunately people have finite lifespans, and by the time someone hits their mid-thirties - children and other priorities will start pulling you away from working 40-60 hours for subsistence wages. The Risk budget for exploring ideas shrinks.
I'm honestly curious if ChatGPT will help in this regard. Working with ChatGPT on physics/math problems allows one to get immediate knowledgable feedback on any idea - you can even have ChatGPT write out a semi-broken version of complex derivations if you just want to see if anything interesting comes out of the idea.
I'm
- robertlagrant 4y agoI'd be curious to know what it's like in big tech firms who are doing quantum computing. Is that well-funded cutting edge science? Or if not, what is it like?
- jobs_throwaway 4y agoI'm more bullish on something like UBI advancing the cause of particle physics. If there was 0 risk that your family was going to go hungry/become homeless, it'd be easier to stomach the many years of training required to get to the forefront of physics.
- atlantic 4y agoIt would be much easier to make the case for UBI for particle physicists than for the general population.
- jerf 4y agoThat's not "UBI" though. That's some funding mechanism that allows for more broad fishing without having to call every shot in advance, and penalizing people who then don't find what they were hoping to find. Though at this point, which is easier between "societal UBI" and "changing how science is funded" is anybody's guess. Nominally, it ought to be the latter, but it may actually be more entrenched....
- jobs_throwaway 4y agoUBI might not be the precise moniker, but I think a UBI that increases with education is a potentially interesting idea. Though I'd worry that we'd get a lot more literature and gender studies PHDs rather than math and physics.
- bigbillheck 4y agoThe world needs more humanities scholars as well.
- SuoDuanDao 4y agoWhat have humanities scholars ever done for me? Granted, some have written books I like, but is there any evidence they wouldn't have done that absent said scholarship?
- jobs_throwaway 4y agoAbsolutely. I'd be in favor of some sort of graduated UBI for students/postgrads working on fields with broad benefits for humanity. $70k/yr or whatever for physics PHDs to be able to research whatever they fancy in the field would probably be an enormous boon to the economy in the long run. You'd have to build in some safeguards from abuse so that it doesn't turn into essentially a pension, but I think you could figure that bit out.
- jerf 4y agoChatGPT qua ChatGPT can never help with the most advanced physics research. By its very nature, it will tend to pull you back to consensus, which is exactly what you don't want it to do in this case. It'll just sit there and plausibly babble about physics. It is super unsuitable for understanding such a mathematical domain that is so unlike language grammar. An AI that uses transformers for language understanding and then adds something more like our frontal cortex might be useful, through.
- ALittleLight 4y agoI think it could, if you could rely on it. Yesterday I was reading an economics paper with a ChatGPT tab open. I was giving it my summary and asking questions about terms and how to interpret them. If what ChatGPT replied with was true, then it was very helpful in getting me to understand the paper and the implications. I imagine the same would hold with physics. Part of research is reading and understanding other papers. If ChatGPT made you 10x better at that, it would help your research.
- jerf 4y agoThe unreliability is fundamental to the nature of GPT, though. It can't not confabulate. You can't know when it is and is not right short of actually doing the work in the first place. The things that will be built on GPT, you might be able to trust. But a language model can't be checking your particle physics work. A language model barely does simple arithmetic correctly. This is not some sort of temporary aberration; even human language models can't do arithmetic correctly. We do that elsewhere.
- lumost 4y agoAnecdotally, chatGPT is much better at higher mathematics and algebra than it is at arithmetic. It’s probably at about the same level in this area as it is for coding related tasks.