I think a lot of people, including futurists, set their standards for “crazy sci-fi future” far too low.
Consider the Industrial Revolution. Take the median person from before it started and after it was over: the daily labor of the latter would be unrecognizable to the former. Their house might look quite different, and the experience of keeping it certainly would be. You would be able to tell the two lived in different worlds just by taking a walk around the block, or looking in their pantry. I don’t know my history well enough to delineate this by exact dates, but if you compare a life in 1780 to its equivalent in 1880 you’d surely see what I mean.
Now imagine all that happened in the span of two years. People would not be unimpressed and disappointed! People would be freaking the fuck out! Imagine taking a five minute walk and seeing your home surrounded by artifacts a person in 2020 wouldn’t even recognize; imagine the best meal you’ve ever had—not just the ingredients but the taste quality too—becoming commonplace. Imagine this happening all at once, too fast to acclimate, at the same time as everyone you know is simultaneously reorganizing their daily routine around the new changes.
I’m not claiming this will or won’t happen in 3-10 years, only that if it did, it would be a really crazy sci-fi future, and I think people would react appropriately. Nothing AI has done to date is really comparable. If you’re a software engineer your job maybe looks a little different now, but I’m a game developer (adjacent at least?) and AI hasn’t affected my daily labors all that much, even though I use it pretty often. I do the same things, I just ask Claude to help with some of them. The math progress is exciting, but I don’t understand most of it, and it doesn’t affect my material existence at all; for me it’s no different than hearing that NASA discovered something cool about Mars, or whatever. And this is as someone with prior interest in the field: I know plenty of people in different bubbles who have never actually had a conversation with an LLM, even now.
Personally, I was pretty impressed and excited by LLM progress from 2020-2024, then kinda disappointed/bummed out about the period after that, though I’m hesitantly considering becoming impressed and excited again given recent events. Most of my impressed/excitedness before 2024 was “within the hobby”, in the sense of “wow, this thing is so much smarter than GPT-3!”, rather than “wow, this thing is restructuring the very bedrock of my existence!” I wonder how much the present-day sense of crazy-sci-fi-ness is predictive (the HuggingFace incident isn’t very important on its own but implies very important things could happen very soon), and how much is looking at things from “within the hobby”, in the sense that they’d only matter to you if you have an existing interest in AI.
I think a lot of people, including futurists, set their standards for “crazy sci-fi future” far too low.
Consider the Industrial Revolution. Take the median person from before it started and after it was over: the daily labor of the latter would be unrecognizable to the former. Their house might look quite different, and the experience of keeping it certainly would be. You would be able to tell the two lived in different worlds just by taking a walk around the block, or looking in their pantry. I don’t know my history well enough to delineate this by exact dates, but if you compare a life in 1780 to its equivalent in 1880 you’d surely see what I mean.
Now imagine all that happened in the span of two years. People would not be unimpressed and disappointed! People would be freaking the fuck out! Imagine taking a five minute walk and seeing your home surrounded by artifacts a person in 2020 wouldn’t even recognize; imagine the best meal you’ve ever had—not just the ingredients but the taste quality too—becoming commonplace. Imagine this happening all at once, too fast to acclimate, at the same time as everyone you know is simultaneously reorganizing their daily routine around the new changes.
I’m not claiming this will or won’t happen in 3-10 years, only that if it did, it would be a really crazy sci-fi future, and I think people would react appropriately. Nothing AI has done to date is really comparable. If you’re a software engineer your job maybe looks a little different now, but I’m a game developer (adjacent at least?) and AI hasn’t affected my daily labors all that much, even though I use it pretty often. I do the same things, I just ask Claude to help with some of them. The math progress is exciting, but I don’t understand most of it, and it doesn’t affect my material existence at all; for me it’s no different than hearing that NASA discovered something cool about Mars, or whatever. And this is as someone with prior interest in the field: I know plenty of people in different bubbles who have never actually had a conversation with an LLM, even now.
Personally, I was pretty impressed and excited by LLM progress from 2020-2024, then kinda disappointed/bummed out about the period after that, though I’m hesitantly considering becoming impressed and excited again given recent events. Most of my impressed/excitedness before 2024 was “within the hobby”, in the sense of “wow, this thing is so much smarter than GPT-3!”, rather than “wow, this thing is restructuring the very bedrock of my existence!” I wonder how much the present-day sense of crazy-sci-fi-ness is predictive (the HuggingFace incident isn’t very important on its own but implies very important things could happen very soon), and how much is looking at things from “within the hobby”, in the sense that they’d only matter to you if you have an existing interest in AI.