Stephen Wolfram Livestreams

Future of Science & Technology Q&A (35 videos)

Biweekly ask-me-anything about the future of science & technology

icon

New episode streaming Friday, March 14, at 3:30pm ET. Submit your questions

Future of Science & Technology Q&A:
Future of Science and Technology Q&A (March 29, 2024) »

1 hour 8 minutes

✕ close
Regarding the April 8 eclipse, what should we know about the weather? What if it's cloudy or raining? What will the future of predicting eclipses look like? Will we be able to move the Moon to keep eclipses? The Moon gets further away because of friction in the oceans from tides, so could we play with that mechanic? I've seen a couple articles about attempting to build trains on the Moon. Would that be useful? Should we expect really unusual emergent properties from LLMs as we scale up the data more and more? ​​​​It seems erroneous to think it's a "more data" question; it's the opposite—"less data" (more generalization/transferability and efficiency) via integration of a "symbolic" framework/accelerator. Human senses capture an absolutely tiny sliver of what's going on. Would additional (strong) senses lead to bigger brains or more languages? Is human consciousness one-dimensional? Perhaps it's not so surprising that training AI on games is useful, when humans learn through play in early life. There is an interesting article about how the requirements of ChatGPT-7 would be massive, like 5 Gorge Dams or half the world's computers, and there isn't enough training data. What does it say about logic (and set/type theories) as a so-called formal language that LLMs can't just trivially do logic and reasoning? What's the best way to pass computational representation to language models, plain text or embeddings? View Less »
View More »
Future of Science & Technology Q&A:
Future of Science and Technology Q&A (March 22, 2024) »

1 hour 6 minutes

✕ close
Do you have any plans, and when can we expect to see a S. Wolfram AI chatbot with voice? Do you think each AI iteration of the person would be similar at the start and then diverge in personality/intelligence as they continued to "live" and develop? ​​​​Are you comfortable with the average quality and correctness of AI-generated answers and commentary? ​​​​Due to the success of nature-inspired computing, I am really wondering if like our best bet is creating full-on human replicas, meaning similar learning experiences/processes... I like the architecture of having many bots with like a base that confer and like upload their findings to a global knowledgebase, then disperse on like new assignments or what have you after returning. ​​​​Do you think human brains compress data in a lossy way, and will future AI brains also have to use lossy compression methods to be more human like? Or would AI perfect memory be more desirable? ​​​​Could an element printer theoretically work, e.g. one feeds it with carbon atoms and it prints out an arbitrary element? ​​Can you explain the quantum LLMs idea, and what advantage exists in applying multi-computation, if any, to LLMs? What would an AI look like that is rewarded based on questioning rather than answering? ​​​​What do you think will happen when we understand prime numbers to their fullest? And when we can translate this knowledge to AI? View Less »
View More »
Future of Science & Technology Q&A:
Future of Science and Technology Q&A (March 8, 2024) »

1 hour 32 minutes

✕ close
Could mind uploading be achieved 20–40 years from now? Memory in human brains is like RAM: it's gone the second you shut the human off. What about copyright for bots based on real people? What are the legal implications? If we could upload our brain to something else, would it be feasible to upload data to our brains or upgrade parts similar to that of upgrading one's computer? What kind of compression algorithm is used on our thoughts? Would it be possible to read brainwaves using AI? ​​​​​​What is the "artificial" part of AI? As far as I can see, LLMs are a breakthrough in the study of intelligence itself. Will technology have an effect on human evolution and ultimately change our physical bodies in the future, i.e. such as our eyes becoming optimized for looking at screens all day, like a built-in blue-light filter, or prioritizing finger shapes to better type on a keyboard or hold a phone... What do you imagine the future of communication to be? We've had spoken word, mailed letters, telegrams, phone calls, emails and now texts. What's next? ​​I need an AI version of Socrates to speak with about being. Are there limitations to using genetic data to understand language-related traits? If so, how might these limitations affect the accuracy or applicability of language models? ​​Do LLMs work well in Egyptian hieroglyphic concepts? Do you regard math as a "language"? If so, would you suggest "talking" math out loud with a child? Stephen mentioned in a recent podcast that certain subject matters in certain academic fields are in a position to be combined with computation and haven't been so far—and stated that it was low-hanging fruit if one was to do it. I was wondering if Stephen could give some examples of these "low-hanging fruits." View Less »
View More »
Future of Science & Technology Q&A:
Future of Science and Technology Q&A (February 23, 2024) »

54 minutes

✕ close
What is the future of business? What's the future of contracting and procurement? Can AI and crypto indeed create "smart contracts," or is that a utopian vision? How do you see universities keeping up with AI? It seems that universities will even more become a preselection for corporations. What are your thoughts on it? What is your opinion on terraforming the Moon? Having corporations with the means (money) as well as nations that have the forethought to have their own supercomputers, how do you see the inevitable increase in the gap between the ultra–high-tech nations/corporations and the rest of the developing world? View Less »
View More »
Future of Science & Technology Q&A:
Future of Science and Technology Q&A (February 2, 2024) »

1 hour 25 minutes

✕ close
Do you think the plateau of LLMs will be at the level of understanding language? Once it does plateau, it will start to taper off and we'll need to use a different technique. How do you know that our brains aren't already using compression? The AI could make far more precise classifications than we make, leading to new words and their very precise and peculiar meanings. What seems more important to me is just the speed of the medium that AI has and how quickly it can process all of the interrelations, even when it's just forty thousand concepts, let alone larger context. Which language is best at this compression? Is there room for significant advancements in mathematical notations? A single biological neuron has been shown to be able to perform an exclusive OR operation. A current simple artificial neuron can't do that. Do you think we are underestimating the power of biological neurons compared to artificial neurons? Did the neural nets that you were playing with learn after their minds were blown? ​​What is your view on emergent phenomena? Do you think it is possible to predict all the emergent abilities AI can possibly obtain? Will humanoid robots usher in an age of abundance? Can a higher intelligence ever transcend computational irreducibility? Why does pessimism toward future technology exist? My experience is that most people's default opinion on future tech is "Let's not do this." Can we calculate how captured wind energy will influence the weather in the future if we become fully independent from fossil fuels? What is there to say about the future of food science? Will someone be able to recreate Willy Wonka's creations? Most of the energy in wind is not on the ground, it's over oceans and up in the sky. ​​Could one devise a "computation engine" analogous to a thermodynamic engine, where useful work is a consequence of information processing? Can AI solve the taste problem as it did the protein-folding problem? If we could run our evolution over and over again, maybe we could see ourselves in some different shapes, floating in space with different ways of communication. For example, we could exchange thoughts telepathically, or maybe we could influence objects without physical contact. What are your thoughts on this? Would it be easier to genetically modify kiwi instead? View Less »
View More »
Future of Science & Technology Q&A:
Future of Science and Technology Q&A (January 19, 2024) »

1 hour 31 minutes

✕ close
Can AI be swallowed by more advanced AI by feeding it via "virtual" input? The motivation could be the increased efficiency of "larger" AI. Do you think anyone will solve the Riemann hypothesis in your lifetime? ​​​Aren't there thousands upon thousand of written papers that assume the Riemann hypothesis is true? Will AIs be the ones to explore space? ​It's ​4.37 light years to Alpha Centauri. We should harness light/light waves so that we can take a picture of the planets there, then bring them back and produce pictures in 8+ years. Do you think Ray Kurzweil's longevity predictions are likely to happen in our own lifetime? Would it make sense for alien AI civilizations to broadcast radio or other signals with information on how to build such AI as a way to propagate in the universe faster than material space travel? People automatically assume that AI will create this type of step function upward in everything in the world or in the economy. This may not be the case due to diminishing returns. View Less »
View More »