Deep Research is a tool to find researched
Deep Research is a tool to find researched
"Deep Research is a tool to find something that someone somewhere on the planet has already researched. It is not a tool to do research that no one on earth has done before".
fuku0185 "The biggest change that has occurred in the world of Shogi due to the evolution of IT and the Internet is that a highway has been laid at once for Shogi to become stronger. But there is a huge traffic jam at the end of the highway" Professional Shogi player Zenji Hanyu. kazumamori_py It would be nice to see a world where there are more incentives in place to research and disseminate what "no one has done yet". Currently, if the project is not brought to the point of successful commercialization, in most cases, no remuneration will be received.
Once ASI is realized, the search for knowledge will probably advance dramatically if we live in a world where AI judges the value of every discovery and appropriate rewards are offered immediately. (Although it will not necessarily be humans who conduct the search for knowledge.)
nishio In the future, if you research something that no one has done yet, describe the knowledge in an easy-to-read form for AI, not for homo sapiens, and deliver it to AI, AI will pay you a royalty every time it uses that knowledge. AI will pay you a royalty every time it uses the knowledge. nishio OpenAI is releasing "GPT Store" on 2024-01-10. I think the reason why not many resources are being allocated to boosting this now is that this is a mechanism that will become more valuable after it becomes a de facto infrastructure, whereas competitors such as Anthropic are moving too fast to de facto it yet, so it is not a priority. Deep Research allows us to quickly learn the "limits of human knowledge".
And beyond that, the significance of the doctorate is to reveal "what mankind does not yet know," and I believe this will become even more important in the future.
https://gyazo.com/e0de63c2ee83abf30ac951db3cdde392
_daichikonno However, I believe that many of the new discoveries are "new combinations of existing knowledge" and many areas will eventually be able to be done by AI alone. (Some are already becoming.)
Just as sports are "physical games" in the modern era, I believe that research will become an "intellectual game" for humans in the mid- to long-term.
nishio I agree with you 70%, but we disagree on the details. More people will be engaged in "jobs that expand the limits of human knowledge that PhDs used to do" and the importance of "PhD education" to do those jobs will increase, but the value of "existing PhDs" will decrease if the number of people meeting the criteria for "existing PhDs" increases tenfold with the help of AI. I think again about what I wanted to say in my original post,
"The value of acquiring scientific research etiquette will increase in the future."
This was the case.
(Nevertheless, we believe that the time will eventually come when scientific research will be completed solely by AI, and we expect that scientific research for humans will become an "intellectual game" of entertainment.
I believe that learning the manners of scientific research is one good investment for the future, as it is one way to live happily in such a future.)
rkmt PhDs and post-secondary education institutions, incl. universities, colleges, etc. are typical " guild", so I am interested to see what happens to those guilds when the social transformation caused by AI is on the scale of an industrial revolution. It may decline when authority and reality no longer match. During the Industrial Revolution, guild organizations from the Middle Ages were reluctant to innovate and eventually disappeared. On the other hand, modern universities, especially those that provide advanced education in science and technology, were born as a result of the Industrial Revolution, and such education and research are now emphasized even in traditional schools. What will happen to universities in the 21st century? rkmt If the purpose is fixed in guilds, they cannot adapt to the changing times. The type of universities sponsored by monasteries could not respond to the industrial revolution, but Cambridge and Oxford (which had been mainly theological and philosophical classics) adapted to the times and became important in the sciences (although there was also Newton to begin with, so it was not just theology). rkmt Cambridge and Oxford were a collection of colleges decentralized, which may have allowed them to have the flexibility to respond to changing times (I asked Dr. CHAT if he He agreed). ---
This page is auto-translated from /nishio/Deep Researchはリサーチ済みのものを見つけるツール using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.