Hi Jose, Regarding finding a PhD program in which to study AGI ...
On Sun, Apr 25, 2021 at 10:31 PM Jose Ignacio Rodriguez-Labra < [email protected]> wrote: > Thank you for the response, Linas. > > I thought that academia was the place to go to work on cutting-edge ideas > and pushing boundaries :( > > However to be honest I do see what you're saying. I've personally spoken > to a couple of experienced professors who've done research on machine > learning and neural networks, but they seem like they don't believe in AGI, > or that it is in a ridiculously away future. I can see their disbelief in > their eyes. > > But I'm sure that in history there have been technologies that no one had > believed in before their development, how did those come to fruition? What > does the history manual recommend I do? Is AGI development from my basement > in my free time the best I can do? How did you end up in the position that > you are that allowed you to gain your expertise in AGI? Thank you. > Those are good questions. To answer them, let me offer a few stories. A long time ago, when I was in grad school, I became very interested in fractals, "negative entropy", and "chaos". I thought these would be fascinating things to study, but was warned against it. I was told that only cranks and crackpots study such things and I would never be able to get a job. This was, of course, shortly before "chaos theory" exploded on the scene. Eventually, there were best-seller books written about it! (In the end, I did not study chaos; I feel like I missed my chance: I was in the right place at the right time and did the wrong thing.) Academics are a conservative bunch, because they have to write grant proposals that do not get rejected. No grant -> no money -> no job/tenure, if you're young, can't do research, can't pay for grad students, can't pay for lab equipment, computer time. The guys in the machine shop who drill pieces of metal for your experiment - they need to be paid. So you don't want to write grant proposals that sound crazy and are likely to be rejected. This makes academics stick fairly closely to the mainstream; they are risk-averse. I did keep whining about chaos, and kept gluing pictures of fractals on the wall. Someone suggested that I talk to Phil Anderson -- he was, by that point, a rather famous and established heretical thinker. However, he was in California, I was in New York, and this was before email. I would have to write a letter, or call on the phone. I was easily intimidated .. and socially awkward. I didn't know how to do that. (I should have done it). I was also sent to talk to Per Bak, a bit before he got famous. He was at Brookhaven; I was almost done with my PhD, I was silly and did not know what to talk about, and I did not know how to ask for a job or how to ask for advice (!!) I knew math and physics, I did not have "ordinary common-sense people skills". Alas. The point of my stories is for you not to repeat them. There almost surely are some AGI thinkers embedded in academia, but I do not know who they are. It is much more valuable and important for you to find them; studying AGI in your basement is a formulas for intellectual suicide. Standard career advice: it is not what you know, it's who you know. Here is an idea: someone recently pointed me at a paper by Geoffrey Hinton "How to represent part-whole hierarchies in a neural network": https://arxiv.org/abs/2102.12627 -- I have not read it yet, but it looks right up my alley. You should raise your hand and write to him, and say something like "hey me! I want to do that!" -- or at least, ask him what you asked me -- advice on getting a PhD. If I may be so bold, I will bcc him on this (bcc, so as not to spam a public mailing list with a private email address.) So here's the deal: I've been working on the intersection of neural nets and symbolic AI since -- I dunno -- 2014? Earlier? I've been laboring in (what feels like) complete obscurity, unable to get anyone interested or excited. Asking me for help is pointless: I have no connections; I don't know of anyone, besides Ben, who is interested in this topic. By contrast: Geoffrey Hinton has a Wikipedia article about him. He has connections both to a respectable university and to Google. That's hard to beat. He may not personally know anyone interested in AGI, but he might be able to point you in the right direction. My strongest advice to you is to keep asking around, through your network of people who know other people who might know, and pose them this same question. I can give you a specific curriculum to study, but this is not the most important thing, right now. Pester me later, as the mood hits. -- linas > On Sunday, April 25, 2021 at 6:52:30 PM UTC-4 linas wrote: > >> I don't know. Let us know if you find out. My general impression is that >> academia is still actively hostile to the idea of AGI. I've never had a >> conversation on the topic that went well. >> >> -- Linas >> >> On Sun, Apr 25, 2021 at 11:49 AM Jose Ignacio Rodriguez-Labra < >> [email protected]> wrote: >> >>> I want to get serious in the development of AGI, and I think doing a >>> Ph.D. might be best for me. I will graduate with my master's this Fall >>> 2021. Do you guys have any recommendations for universities, research >>> institutes, or programs in the US or even international? Thank you. >>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "opencog" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/opencog/d0efa315-7d7d-4fad-8ba3-33432981f0e5n%40googlegroups.com >>> <https://groups.google.com/d/msgid/opencog/d0efa315-7d7d-4fad-8ba3-33432981f0e5n%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> >> >> >> -- >> Patrick: Are they laughing at us? >> Sponge Bob: No, Patrick, they are laughing next to us. >> >> >> -- > You received this message because you are subscribed to the Google Groups > "opencog" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/opencog/e3426270-cfb1-4df0-82ac-fede670ab5d8n%40googlegroups.com > <https://groups.google.com/d/msgid/opencog/e3426270-cfb1-4df0-82ac-fede670ab5d8n%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- Patrick: Are they laughing at us? Sponge Bob: No, Patrick, they are laughing next to us. -- You received this message because you are subscribed to the Google Groups "opencog" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/opencog/CAHrUA34f1KxcQTyV3uSvtVzQz316rhqizR8jp835be%3Dv9C121A%40mail.gmail.com.
