Stephen Wolfram answers questions from his viewers about the future of science and technology as part of an unscripted livestream series, also available on YouTube here: https://wolfr.am/youtube-sw-qa

Questions include: Can you comment on the future of LLMs being running in the cloud vs. being run on one's local machine? - Does the NANOGrav discovery spark ideas for experimental validation of the Physics Project? - Can you discuss the next evolution for AI models? So far we have: language models, image – text (classifiers), text – image (generators), etc. - What can be said for training multimodal AI models? - Do you think that we have reached a point of singularity such that any child born from today onward will never be able to surpass AI at any intellectual task, i.e. are we the last "useful" generation? - Is VR the future of UIs? - Given the two contrasting scenarios of a "Pink Plasma Heaven," where artificial general intelligence optimally solves problems for all sentient life, and a "Matrix Hell," where AI exploits humans as energy sources, how can we establish a guiding framework to navigate between these extremes? - To what degree do you think LLMs provide us with insights on the internal workings of our brain? Do you think there will be more lessons to learn from the structure of the human brain when designing the next generation of LLMs? - Does the spread of LLMs incentivize scientists (and humans in general) to become more deeply specialized (to "out-compete" LLMs in a narrow domain) or to become more broadly spread (in order to creatively generate connections between apparently remote domains)? - Will it be possible to use LLMs to achieve world peace? Or if world peace isn't big enough, can we beam LLM chats into outer space to try and get universal peace? - What do you think of power laws? What do you think are some good entry points for explaining the principles behind power laws? - What do you think of the future of AI in video games? They can be used to control the actions and dialog of NPCs, the design of the game's world and even the design of assets on the fly using little data. Video game assets can take up a lot of data, and if we could use AI to generate assets on the fly using a smaller amount of data, we could cut down on the download size of games as well as the effort needed to make assets. - How will we be able, in the future, to tell what we're seeing on screen isn't AI generated? Anything we could do today? (I think you might be a bot.) - Thinking in terms of inter-concept space, do you think there is an approach to using technology to develop a way in which we may better understand or gain experience to bridge the gap of inter-concept space between what we know and what we don't know? - When will this statement, "I think you might be a bot", be a compliment, rather than a criticism or an insult?

Future of Science and Technology Q&A: Live from the Wolfram Summer School (July 7, 2023)

Stephen Wolfram is the creator of Mathematica, Wolfram|Alpha and the Wolfram Language; the author of A New Kind of Science; and the founder and CEO of Wolfram Research. Over the course of nearly four decades, he has been a pioneer in the development and application of computational thinking—and has been responsible for many discoveries, inventions and innovations in science, technology and business.

On his podcast, Stephen discusses topics ranging from the history of science to the future of civilization and ethics of AI.

Future of Science and Technology Q&A: Live from the Wolfram Summer School (July 7, 2023)

The Stephen Wolfram Podcast

Related tracks