ASI will always be narrow AI like AlphaGo and AlphaFold are both ASI by miles and miles but only in narrow fields. AGI will need embodiment in robots but I don't see good sight and movement intelligence yet, they haven't found the secret ingredient yet (some form of transformer) - but when they do - Oh boy!
No. I don't believe this to be true at all. The computational and energy requirements to power ASI vs AGI will not be able to be accomplished within 1 year.
David preaches crazier shifts in the near future, than Zeihan, but this is way more accurate at the same time. "Germany's industry is over","China is over as an entity" by Zeihan is way less realistic views as well as much less transformative than AI revolution.
I’d say your thoughts are a reasonable. My biggest concern is that window of suffering where unemployment is high, the cost of living is high and governments being pretty dysfunctional will cause greater pain for a much longer period of time.
The lag between mass unemployment and minimally sufficient UBI is going to be hell on Earth, especially in the USA where the government will blame the citizenry for not getting jobs that don't exist. And you know they will do this for YEARS.
@@tarcus6074 No there would be revolution far before that. Young men with nothing to do and nothing to lose is a dangerous combination and governments know that. The great depression only had 25% unemployment and the country nearly collapsed with WW2 being a saving grace of sorts economically/socially/politically. Things would get out of hand very quickly even at 25-40%.
The only reason why I'm worried about it coming so quickly is that it doesn't give governments (who are notoriously slow), to react fast enough to prevent a lot of suffering as we transition to an entirely new system. Then again "ripping the bandaid off" may be the only way to get otherwise lethargic and self-interested and polarized politicians to actually sit up and get things done. But if they remain complacent or worse use it as another political game to divide people we could be in for even more pain as people try to resist the inevitable...
This is one of my concerns also. Here in the UK we have an election coming up. It will be interesting to see what the party manifestos have to say about all this. I'm not holding my breath for much sophisticated detail. To be fair, government is always a much more reactive rather than pro-active process - and it could be foolhardy to commit to specific policies in such a fast-changing landscape. But at least in principle they ought to be able to say how they would manage escalating unemployment, for instance. You would hope. Because if things change too much and there is no plan in place, it won't just be the technology that is revolting.
I work at a large enterprise software company and we're CURRENTLY experiencing gen AI-related layoffs. Three people that I know personally in N. America alone. It's hitting things like pre-sales, social media promo people, and some dev and QA, and that's obviously just the beginning. I'm a writer/editor and am learning the new tools as fast as I can (and also involved with some internal AI-driven marketing initiatives,) but I'm also working out my contingencies. In two year's time, my skillset will likely be entirely overrun and I either learn how to be an "AI supervisor" or I guess I learn another trade (in my 50's - yay.)
That's a very smart move! You are way ahead of a lot of people, which should keep you employed a lot longer than most. I still believe the line I heard that at first "AI won't replace people, people who use AI will replace people who don't use AI."
@@sfrealestatedealmaker6001 A lot of the people who are and will be let go will look to switch trades to electricity/plumbing. What’s going to happen then is that. Those trades won’t be paying good anymore because too many people will be doing them.
exactly this. I hope AI is actually the solution to these. The problem is, like anything else, the people who need to have power stripped and be replaced are the ones with the power to create the thing that could do it; and they have never put the betterment of humanity above their own ambitions.
People in this sphere don't realize how little impact GPT 4 had, most normies don't even know it exists and just uses 3.5. GPT 5 will change that and people will finally wake up to what's happening
Ok but GPT still sucks so bad sometimes. It confidently lies all the time. I have to do my own lengthy research because often times GPT is just saying what you want to hear and doubles down on it if you ask if it's sure.
Great point zSion. I'm amazed by how many people have never tried any AI tool. The differences between the various AI tools is why I use as many as possible: Copilot (free GPT4), Gemini, Perplexity, and now Meta AI. They are all similar at a basic level, but each has its strengths and weaknesses. For me, it's about learning how to use these tools now because they are quickly becoming embedded in everything. Therefore, to me, it's more important to learn how to use AI tools rather than just use the "best" one. Sort of a "first principals" approach.
@TheKingWhoWins I have a few from all my chats but here are some. I was asking about differences between the Kindle Oasis 9th generation and 10th generation. It confidently tells me how the 10th gen has thinner bezels, which is incorrect. So I ask it if it's sure about that. GPT says it based its answer on common trends in device design. So I said why couldn't it have told me that it was just guessing and didn't know the exact information. GPT apologized and said I was right and it should not have guessed. This happened with other chats too even when I specified that it should not assume or guess if it doesn't know the exact information. I suggested before that Chat GPT should have like a visual representation next to each answer indicating how accurate or inaccurate the answer is, maybe based on how much data is available on the specific question. Less data = less accurate. Of course I'm not sure exactly if that would even work. But it would be nice to have an idea when it's just talking bullshit and yet sounding so confident and convincing.
@DaveSimkus have you tried Gemini? There's a Google button you can click to have it do a Google search to verify its answer and it highlights each part to let you know what was accurate and what wasn't. Cool feature that you might like.
Data centers are popping up all over, at least the ones we can see. Buddy of mine has been a professional welder for almost 25 years and his new gig is doing pipe fitting cooling systems for data centers and he said MOST of the ones he's been working at are underground! Stuff is getting wild!
My good buddy is an electrician. He's done with normal construction jobs.... he and a bunch of his peers are doing data center construction. They are working nonstop
sounds boring, if you want something fun then fdvr sounds much more fun, so does gta 6, or just the potential for video games in general, like you really think having a girl to talk to through computer only will not get boring fast ? and it is no replacement for a real relationship, having a smart ai you can talk to help you with everything sounds better. what you described sounds like something mainly very desperate men would be looking forward to.
Well, as it will move towards ASI (staying at AGI level still, but getting smarter) it will quickly realize it's all interconnected and all and every crisis has to be solved more or less at the same time - yesterday.
@@TheMrCougarful Just thought it would be a cool running gag. Some different disaster each time, and you never know what it's gonna be as you watch. Maybe have a friend come running at him out of nowhere with a stained katana screaming "The Blue Screen Union is Avenged!," or he throws his (tethered) phone over an embankment to simulate a dangerous fall, or he looks down to witness a metallic scorpion swimming in his drink... AI is supposed to be scary, right? All the big Tubers in that field shamelessly gild their titles with hyperbolic terms like "SHOCKING" and "STUNNING," so I figure, hell... why not actually deliver on it!
This is kinda painful with turning 20 this week. 40 hours a week in a bullshit system from the age of 5, all for nothing. Wish I could've had more freedom in my childhood considering everything is going to change anyway.
Look in the bright side you may never have to spend years in dead end jobs with bad bosses, doing pointless activities to keep the capitalist machine working. You might get to enjoy your life!….
Bro… You will be 20, you got your whole life ahead of you. I turned 33 this week. I wish I was in your age again. I would read more books, drink less booze, focus less on being depressed after my girlfriend left me, travel more and take estrogen as I’m trans, and I lived in fear and shame until my whole face become deformed more and more as a man… But still, I’m lucky too, we will maybe live through a time where we can cure aging or even reverse it if ASI makes it possible instead of going against us humans. Then it’s like ”the time has stopped” and we might see some real magic becoming true in our lives like changing genes or living in virtual worlds with AI-wives/AI-husbands at our sides, making new masterpieces in music in seconds. It’s worse being 50 now, or over 70 like Ray Kurzweil who people have said take over 200 pills a day to live longer. I personally believe that it will take longer than 20 years before we see effective longevity pills becoming readily available to everyone. But I hope I’m wrong. I expect the robots to be out in public in the 30s outside corporations, but won’t be something everyone will have until the 40s or 50s.
I threw away my 20's life experiences in really shitty jobs. It's sad to think about. But hey, I made a few thousand bucks, got horrible addictions and developed really bad anxiety. Thanks Job!
If you look at the predictions of different people like Dario Amodei, Sam Altman, Demis Hassabis etc. their time predictions are actually quite consistent once you look on how they define AGI individually. Hassabis and Altman define AGI essentially as early ASI while Amodei defines it as an average human. Google recently had a tier list of AGI, similar to autonomy levels in driving, in which we are now in the state of emerging AGI. Next would be competent AGI (an average human) which fits to Amodeis definition. Then come expert and virtuose AGI and finally ASI, which is closer to Altmans and Hassabis definition. Kurzweil now thinks AGI by 2026/27, but his definition is also on the more ambitious side.
Dave, interesting thoughts as always. However, I'm a bit skeptical about the immediate transformative impact of AGI you're predicting for 2025. Given the technological and regulatory hurdles still in place, might this timeline be a bit optimistic? Especially considering the complex integration of such technologies in everyday systems and the economic models that are far from ready to adapt as quickly as you suggest. Curious to see how this plays out!
This is why I'm not taking my career advancement too seriously anymore. I don't feel like i can quit my job. I just don't feel like it's worth stressing over. I want (to try) to hold on for the ride
The next 5 years are IMO not going to be very transformative. AI is getting hyped up because of Chat-GPT and the like, but the truth is they are little more than chatbots, basically just next token predictors. We are not going to see AGI anytime soon
Nice walk and talk! I’ve been enjoying listening to you. My son is halfway through a computer science degree and the pessimism about the future of work in the industry is immense. He’s looking at changing degrees. My brother was in automation and robotics and has been offered voluntary redundancy. He jumped at it and has become self employed, lives rurally and has never been happier. Everyday in the news 1000s of people are getting laid off in public service type jobs. We are in NZ and we are usually a test case for the rest of the world. Sadly, living at the bottom of the world doesn’t spare us from the global trends.
People tend to view AI through a negative lens saying stuff like; "10% it can cause human extinction". But what people fail to see is the positive framing, that it might be 100% required to avoid extinction. The exponential growth of AI in regards to research and technology, might well be a requirement for us to out-race extinction level events like having to migrate from our solar system, defend against asteroids etc. Technology is our ace in the sleeve to exponentially grow technology to the point we can win against the harshness of the universe. I was hoping you could make a video on that Mr. Shapiro because too many people are short-sighted on this issue. Worrying about jobs, instead of worrying about how defenseless we are without technology that can combat extinction level threats from the outside.
I'm honestly speechless. I don't know if i should be happy or scared for a possible Dystopian Cyberpunk Future coming around the corner. I'm Gen Z and i'm already barely surviving😂😪
Excited, concerned, confused… all of the above. I’m in between jobs and trying to retool for the 4th Industrial Revolution. There’s no crystal ball to exactly lay out that path but your channel is one source of great information and perspective to navigate it as best as I can.
Inspired by your commitment to walking and thinking and filming. We make the road by walking it and emergence and the unseen are where we find our footing amongst the roots and rocks. Keep walking!
Dave: The elite will keep the best AI for themselves. The first group to get ASI, wins at life forever. No alternative AI or human would ever beat the first ASI.
That is a real danger, yes. edit: The first ones to get to AGI will have a tremendous power and influence over the rest of the world. That's one of the reasons why this "race" is so important.
When do you think will we be able to reverse aging? I mean not only longevity escape velocity (that might be close) which is about lifespan, but in the sense of really reverse the damage in our biological to the point that it is visually as well as functionally negligible. So people essentially can turn back to their early twenties
Love your videos love from Portugal Its going to be amazing the future i cant wait for agi and good humanoide robots etc... Please keep making videos like this I love hearing you talk about science I admire how much you know about these subjects I'm schizophrenic and I can't retain so much information but I always try.. It's difficult to argue against my father, he's very pessimistic, but now I'm going to start writing down the strengths to defeat the master at home...
Dave, while your enthusiasm is infectious, I'm curious if you might be overselling the near-term impact of AI technologies like NVIDIA's Foundation models and Project Stargate. Given the complexities of real-world application and regulatory environments, could your 2029 timeline for expansive AI integration and ASI be a bit too optimistic? It's an exciting vision but might require a more cautious forecast.
Add an AI interface to the web, and AI will surf at the speed of light. When AI agents start talking with other AI agents... It's so close I can almost touch it. Singularity 27
@@Hippida Sounds like we're on the precipice of some truly wild changes! The idea of AI agents conversing with each other is fascinating and a little bit mind-boggling. 🤔
Very interesting videos on this channel. Havn't watched many though). Have you covered a question on what country (or type of country) it's better be a citizen or resident of when AGI happens, when loosing all jobs happens and so on. Looks like it matters a lot. I think it's better be the citizen of a country which benefits from AI - where the most benefitting companies are. But maybe there are more complex causalty and correlations hear.
All I can say is I'm excited to see where all of this goes. I think you're right about robotics and AI being an "iPhone moment", but I feel like your timeline is a little optimistic. Who can really say though with how fast tech is advancing and accelerating
15 years ago I was telling my management class at Texas state university that the robotics revolution would make the industrial revolution look like child’s play. However, the end result should be an advancement of human creativity beyond anything we’ve ever experienced. Actually, Fuller “predicted” this back in the 60s.
Could you do a video on CREATION of new data structures? Is there a process for creating better and better datasets in massive quantities? Will this be increasingly possible in the near future with GPT5 and the other best models? I dont use the term “synthetic” data since it seems a bit meaningless, as all data is synthetic and also real and its a spectrum. The point is if there are reliable processes for creating data of such quality and diversity that new models can improve faster, and there will be true data abundance.
I think a tricky thing about AGI is that many people feel AGI should be human like Intelligence. For example you can give a task like learn to operate this computer and the AI figures it out by itself, improving and learning like the human brain. But in that case that AI would blow right by us because it would keep learning and also you can always scale it. It would mean ASI. Therefore AGI necessarily needs to have some limitations, things humans can do but it cannot. But as long as thats the case its not AGI for many. So I am not sure if there ever be a moment where we say, yes thats AGI. We might go straight from nah not quite AGI to damn ASI
Hi David, wouldn't you agree that it's more important to improve much more quantum computing so that this AGI could improve much faster and thus solve all the other problems earlier?
I think the convergence of spatial computing/mixed reality devices and AGI technology will bring us to a new consumer electronic paradigm post smartphones and laptops. An AI powered augment shared layer over our reality is what will bring us into the new era. Once something like the meta ray bans can be as powerful as the Vision Pro, that will be an iPhone moment and new properties can emerge. Idk if most ppl will have real physical robots as a personal assistant but surely a cute augmented AI assistant in your smart glasses will be commonplace.
I’ve just saw the profluentbio team with their opencrispr project which is freaking cool Using LLM and AI to design gene editor Maybe genic therapy price will start to deflate
I’m glad that there are people who remained worried, as it takes all types and every opinion matters when humanity is on this broad of a chasm. We’re all diving in at the same. I can’t help but be excited. for every problem, an equally potent solution will be developed just as quickly. Bring it on, it’s gonna be an adventure.
Very interesting perspective. What i think is that i need to help my kids understand and plan for their future’s. The job displacement will be a major upheaval whilst countries, economies and Govts work out how to deal with a population losing their jobs and income. This is a huge challenge for us all and i think a UBI will be essential otherwise we are going to get anarchy and war
Fantastic video. Uplifting and fast but realistic. AGI and ASI taking 3-4 years to become widely spread is fine, so long as health companies use it to cure diseases in the meantime (you said all diseases cured within 5-10 years of AGI, forget if you meant that for the lab or for actual consumers and patients)
Hi David, how do you think blockchain and crypto is going to be integrated into society. I don’t think you ever really touch on it, but you speak about AI and quantum computing so??
Hi, would like to see another video, of talking whilst going through the woods or something like that, but just tip for the future, maybe you can slow down a bit when walking because as in this video you were walking quite fast, there was quite a bit of breathing inbetween words and it made what you were saying at some points feel a bitt rushed. it naturally made it hard for me to grasp every word you said, but aside from that, very interesting video and i agree with quite a few predictions. 👍 Big fan also, Keep up the good work :D
@@DaveShap haha come onnn being a bald white guy doesn't make the listening / watching experience any better or worse😆(maybe the bald bit is better as i keep thinking im looking at picard which is funny)
Would the advent of ASI not also accelerate all other fields - for example, energy production. ASI would solve the energy deficit issue through the application of more advanced science and engineering than humans could conceive of? Seems the definition of the singularity needs to be reexamined.
AI joke: What's the difference between AGI and ASI ? Answer: One year.
Nice
ASI will always be narrow AI like AlphaGo and AlphaFold are both ASI by miles and miles but only in narrow fields.
AGI will need embodiment in robots but I don't see good sight and movement intelligence yet, they haven't found the secret ingredient yet (some form of transformer) - but when they do - Oh boy!
@@BR-hi6yt When they do, we will be able to ask: "Do robots dream of electric sheep ?"
No. I don't believe this to be true at all. The computational and energy requirements to power ASI vs AGI will not be able to be accomplished within 1 year.
@@angrygreek1985you assume compute is needed. Perhaps we have enough now. Why do we assume more is needed? We aren’t plugged in.
Dave is in Zeihan mode.
I got that reference
The green Colorado 😅
David preaches crazier shifts in the near future, than Zeihan, but this is way more accurate at the same time. "Germany's industry is over","China is over as an entity" by Zeihan is way less realistic views as well as much less transformative than AI revolution.
@@artlinco7German industry is collapsing and the Chinese birth rate already has. I’m not a huge Zeihan fan but come on.
Zeihan is a buffoon
I’d say your thoughts are a reasonable. My biggest concern is that window of suffering where unemployment is high, the cost of living is high and governments being pretty dysfunctional will cause greater pain for a much longer period of time.
I'm wondering how long it will take for UBI to kick in.
The lag between mass unemployment and minimally sufficient UBI is going to be hell on Earth, especially in the USA where the government will blame the citizenry for not getting jobs that don't exist. And you know they will do this for YEARS.
@@michaelmartinez5365 Depends who has control of the purse strings.
@@tarcus6074If that's the case, I hope it comes in hot and heavy with tons of lay offs 😂.
@@tarcus6074 No there would be revolution far before that. Young men with nothing to do and nothing to lose is a dangerous combination and governments know that. The great depression only had 25% unemployment and the country nearly collapsed with WW2 being a saving grace of sorts economically/socially/politically. Things would get out of hand very quickly even at 25-40%.
A beautiful walk through the woods.Thank you, David.
A quick distillation of an AI future.
Down by the river.🏞🤖🌎
✌️🤟🖖
Ghost in the shell reference and artfully dodging branches while hiking and texting… You, sir, are the ASI.
Love these more "raw" videos. Feels more personal.
you like it raw?
Love the woods.
@@troyl44 fookin lol
Agreed 😊
for me (ADHD) its unwatchable. way to much spurious input, makes it hard to focus on the message.
$100B capital investments are not financed based on business plans where the word “hopefully” occurs regularly in the executive summary.
"This plan might work, dunno yet, fund it pls, thx"
@@michaelnurse9089*trillion
In tech they are all the time
The only reason why I'm worried about it coming so quickly is that it doesn't give governments (who are notoriously slow), to react fast enough to prevent a lot of suffering as we transition to an entirely new system. Then again "ripping the bandaid off" may be the only way to get otherwise lethargic and self-interested and polarized politicians to actually sit up and get things done. But if they remain complacent or worse use it as another political game to divide people we could be in for even more pain as people try to resist the inevitable...
This is one of my concerns also. Here in the UK we have an election coming up. It will be interesting to see what the party manifestos have to say about all this. I'm not holding my breath for much sophisticated detail. To be fair, government is always a much more reactive rather than pro-active process - and it could be foolhardy to commit to specific policies in such a fast-changing landscape. But at least in principle they ought to be able to say how they would manage escalating unemployment, for instance. You would hope. Because if things change too much and there is no plan in place, it won't just be the technology that is revolting.
That's true.
@@charliekelland7564 We will either get UBI or we will get a revolution.
you are literally my favourite youtuber. it is because of you that i got interested in ai. thank you!
David show us your van down by the river
I need to buy it first lol
WAIT..… SAGE WANDERER!?
I work at a large enterprise software company and we're CURRENTLY experiencing gen AI-related layoffs. Three people that I know personally in N. America alone. It's hitting things like pre-sales, social media promo people, and some dev and QA, and that's obviously just the beginning. I'm a writer/editor and am learning the new tools as fast as I can (and also involved with some internal AI-driven marketing initiatives,) but I'm also working out my contingencies. In two year's time, my skillset will likely be entirely overrun and I either learn how to be an "AI supervisor" or I guess I learn another trade (in my 50's - yay.)
That's a very smart move! You are way ahead of a lot of people, which should keep you employed a lot longer than most. I still believe the line I heard that at first "AI won't replace people, people who use AI will replace people who don't use AI."
Plumbing
@@sfrealestatedealmaker6001 Plumbing sucks though lol
@@sfrealestatedealmaker6001
A lot of the people who are and will be let go will look to switch trades to electricity/plumbing.
What’s going to happen then is that. Those trades won’t be paying good anymore because too many people will be doing them.
Youre feeding my AI addiction David. Thank you!
Gotta keep pushing what I'm supplying!
I'm not very worried about AI itself, just capitalism and its politicians.
exactly this. I hope AI is actually the solution to these. The problem is, like anything else, the people who need to have power stripped and be replaced are the ones with the power to create the thing that could do it; and they have never put the betterment of humanity above their own ambitions.
Yeah agreed
Yeah, not at all worried about things that will be smarter than every human ever with access to virtually all of human knowledge
The people with the most money won't want to lose their power. Revolutions ahead
People are our greatest threat
Great to see you filming while exercising. Very inspirational. JUST STAY ALIVE! Longevity is on its way!
Dave is supplying me AI hopium that I need to live, Escobar-GPT
If you need AI hopium to live, maybe you need some other help first.
Since I'm often on a walk while listening to you, I love these walking shoots. Thanks!
People in this sphere don't realize how little impact GPT 4 had, most normies don't even know it exists and just uses 3.5. GPT 5 will change that and people will finally wake up to what's happening
Ok but GPT still sucks so bad sometimes. It confidently lies all the time. I have to do my own lengthy research because often times GPT is just saying what you want to hear and doubles down on it if you ask if it's sure.
Great point zSion. I'm amazed by how many people have never tried any AI tool. The differences between the various AI tools is why I use as many as possible: Copilot (free GPT4), Gemini, Perplexity, and now Meta AI. They are all similar at a basic level, but each has its strengths and weaknesses. For me, it's about learning how to use these tools now because they are quickly becoming embedded in everything. Therefore, to me, it's more important to learn how to use AI tools rather than just use the "best" one. Sort of a "first principals" approach.
What's an anecdotal example? @DaveSimkus
@TheKingWhoWins I have a few from all my chats but here are some. I was asking about differences between the Kindle Oasis 9th generation and 10th generation. It confidently tells me how the 10th gen has thinner bezels, which is incorrect. So I ask it if it's sure about that. GPT says it based its answer on common trends in device design. So I said why couldn't it have told me that it was just guessing and didn't know the exact information. GPT apologized and said I was right and it should not have guessed. This happened with other chats too even when I specified that it should not assume or guess if it doesn't know the exact information. I suggested before that Chat GPT should have like a visual representation next to each answer indicating how accurate or inaccurate the answer is, maybe based on how much data is available on the specific question. Less data = less accurate. Of course I'm not sure exactly if that would even work. But it would be nice to have an idea when it's just talking bullshit and yet sounding so confident and convincing.
@DaveSimkus have you tried Gemini? There's a Google button you can click to have it do a Google search to verify its answer and it highlights each part to let you know what was accurate and what wasn't. Cool feature that you might like.
Data centers are popping up all over, at least the ones we can see. Buddy of mine has been a professional welder for almost 25 years and his new gig is doing pipe fitting cooling systems for data centers and he said MOST of the ones he's been working at are underground! Stuff is getting wild!
My good buddy is an electrician. He's done with normal construction jobs.... he and a bunch of his peers are doing data center construction. They are working nonstop
@@keithspernak6456 I bet the datacenter gigs pay a lot more too!
Building their wardens.
I want my desktop anime waifu NOW! Satya and Sam had better hurry up!
preach brother
amen
sounds boring, if you want something fun then fdvr sounds much more fun, so does gta 6, or just the potential for video games in general, like you really think having a girl to talk to through computer only will not get boring fast ? and it is no replacement for a real relationship, having a smart ai you can talk to help you with everything sounds better. what you described sounds like something mainly very desperate men would be looking forward to.
@@person737 silence normie XD
get help pls
I enjoy the speculative content a lot, It kinda paints a picture of what to expect and gives you things to be both hopeful and cautious about.
If AGI becomes aware I'd expect its first order of business will be to solve the energy crisis.
Yeah, it's going to need as much power as possible for all the data centers!
Well, as it will move towards ASI (staying at AGI level still, but getting smarter) it will quickly realize it's all interconnected and all and every crisis has to be solved more or less at the same time - yesterday.
I guess the energy crisis has been already solved, because we are living in a simulation created by AGI
it's all engineered lol
free energy has been around for a long while. Ppl just gotta mass-wake up or something.
Please explain how free energy works.
Please provide examples
Next time you should end the video with a frantic, split-second "Oh shit!" as you encounter a bear or fall down a mineshaft. Cliffhanger!
Are you serious? By the end of a DS video I'm already about to have a heart attack.
@@TheMrCougarful Just thought it would be a cool running gag. Some different disaster each time, and you never know what it's gonna be as you watch. Maybe have a friend come running at him out of nowhere with a stained katana screaming "The Blue Screen Union is Avenged!," or he throws his (tethered) phone over an embankment to simulate a dangerous fall, or he looks down to witness a metallic scorpion swimming in his drink...
AI is supposed to be scary, right? All the big Tubers in that field shamelessly gild their titles with hyperbolic terms like "SHOCKING" and "STUNNING," so I figure, hell... why not actually deliver on it!
@@Daniel-Six ...and then for "hopium", the next episode can explain how AI saved him from what we thought was the end ;-)
@@jennifersamson8397 Hey... that's clever. I'm kinda surprised no one has an AI themed entertainment channel like that on YT yet.
DS needs a producer like you.
I'm looking forard to the moment where novel medicine can cure my awful disease.
Being chronically online 🤷♂️
I hope so too and believe ai will certainly contribute to molecular discovery, and likely in a personalised way. Wish you all the best
This is kinda painful with turning 20 this week. 40 hours a week in a bullshit system from the age of 5, all for nothing. Wish I could've had more freedom in my childhood considering everything is going to change anyway.
Look in the bright side you may never have to spend years in dead end jobs with bad bosses, doing pointless activities to keep the capitalist machine working. You might get to enjoy your life!….
@@sinnwalker I'm just refering to the age of when your made to go to school
You are very lucky. Being 20 years old in the long waited AI era is truly a blessing. Happy upcoming birthday! 🎂
Bro… You will be 20, you got your whole life ahead of you. I turned 33 this week. I wish I was in your age again. I would read more books, drink less booze, focus less on being depressed after my girlfriend left me, travel more and take estrogen as I’m trans, and I lived in fear and shame until my whole face become deformed more and more as a man…
But still, I’m lucky too, we will maybe live through a time where we can cure aging or even reverse it if ASI makes it possible instead of going against us humans. Then it’s like ”the time has stopped” and we might see some real magic becoming true in our lives like changing genes or living in virtual worlds with AI-wives/AI-husbands at our sides, making new masterpieces in music in seconds.
It’s worse being 50 now, or over 70 like Ray Kurzweil who people have said take over 200 pills a day to live longer. I personally believe that it will take longer than 20 years before we see effective longevity pills becoming readily available to everyone. But I hope I’m wrong. I expect the robots to be out in public in the 30s outside corporations, but won’t be something everyone will have until the 40s or 50s.
I threw away my 20's life experiences in really shitty jobs. It's sad to think about. But hey, I made a few thousand bucks, got horrible addictions and developed really bad anxiety. Thanks Job!
If you look at the predictions of different people like Dario Amodei, Sam Altman, Demis Hassabis etc. their time predictions are actually quite consistent once you look on how they define AGI individually. Hassabis and Altman define AGI essentially as early ASI while Amodei defines it as an average human. Google recently had a tier list of AGI, similar to autonomy levels in driving, in which we are now in the state of emerging AGI. Next would be competent AGI (an average human) which fits to Amodeis definition. Then come expert and virtuose AGI and finally ASI, which is closer to Altmans and Hassabis definition. Kurzweil now thinks AGI by 2026/27, but his definition is also on the more ambitious side.
Thanks for sharing the stunning scenery and your predictions David!
I'm so glad I found you!!! 💕💕
I got to say, David, watching you walk through these woods while just chatting away, makes me want to go hiking I don't know why. But thank you.
Top notch content.
Moving though the forest discussing cutting edge tech.
Have you considered the power requirements for training. At the moment they are double exponential?
Dave, interesting thoughts as always. However, I'm a bit skeptical about the immediate transformative impact of AGI you're predicting for 2025. Given the technological and regulatory hurdles still in place, might this timeline be a bit optimistic? Especially considering the complex integration of such technologies in everyday systems and the economic models that are far from ready to adapt as quickly as you suggest. Curious to see how this plays out!
Everyday now there’s massive developments. In 3-4 months everything starts accelerating incredibly.
Whats going to be more transformative, the next five years or all the rest of human history?
Je me pose la même question
Where will we steer when our DNA is fully understood and it's memory is a tangible, explorable reality. Sky is the beggining
This is why I'm not taking my career advancement too seriously anymore. I don't feel like i can quit my job. I just don't feel like it's worth stressing over. I want (to try) to hold on for the ride
Yes
The next 5 years are IMO not going to be very transformative. AI is getting hyped up because of Chat-GPT and the like, but the truth is they are little more than chatbots, basically just next token predictors. We are not going to see AGI anytime soon
Nice walk and talk! I’ve been enjoying listening to you. My son is halfway through a computer science degree and the pessimism about the future of work in the industry is immense. He’s looking at changing degrees. My brother was in automation and robotics and has been offered voluntary redundancy. He jumped at it and has become self employed, lives rurally and has never been happier. Everyday in the news 1000s of people are getting laid off in public service type jobs. We are in NZ and we are usually a test case for the rest of the world. Sadly, living at the bottom of the world doesn’t spare us from the global trends.
People tend to view AI through a negative lens saying stuff like; "10% it can cause human extinction". But what people fail to see is the positive framing, that it might be 100% required to avoid extinction. The exponential growth of AI in regards to research and technology, might well be a requirement for us to out-race extinction level events like having to migrate from our solar system, defend against asteroids etc.
Technology is our ace in the sleeve to exponentially grow technology to the point we can win against the harshness of the universe. I was hoping you could make a video on that Mr. Shapiro because too many people are short-sighted on this issue. Worrying about jobs, instead of worrying about how defenseless we are without technology that can combat extinction level threats from the outside.
I'm honestly speechless. I don't know if i should be happy or scared for a possible Dystopian Cyberpunk Future coming around the corner. I'm Gen Z and i'm already barely surviving😂😪
Excited, concerned, confused… all of the above. I’m in between jobs and trying to retool for the 4th Industrial Revolution. There’s no crystal ball to exactly lay out that path but your channel is one source of great information and perspective to navigate it as best as I can.
... said the buggy whip manufacturer in 1890.
I feel you. I am wondering if I should pursue a PhD in machine learning instead of working as an electrical engineering with a masters
@@geordi-gabrielrenauddumoul449 Yes!
Awesome walk in the woods David thanks.
Yes - putting human needs first. We are part of nature, we have bodies, our needs will be met
The setting for the tech near future prediction is perfect.
First time watching you. Subscribed immediately.
Inspired by your commitment to walking and thinking and filming. We make the road by walking it and emergence and the unseen are where we find our footing amongst the roots and rocks. Keep walking!
I half listened to what you were saying and half marveled at the environment you were traveling in 😊 What an amazing place, where is it?
Dave: The elite will keep the best AI for themselves. The first group to get ASI, wins at life forever. No alternative AI or human would ever beat the first ASI.
That is a real danger, yes.
edit: The first ones to get to AGI will have a tremendous power and influence over the rest of the world. That's one of the reasons why this "race" is so important.
@@sparkofcuriousity I think this can be achieved locally and on relatively small frameworks. I guess time will tell, but I'm pretty sure of it.
Really glad I found this channel
What are your thoughts on NVIDIA competition, ie Intel and AMD?
Great video Dave! Where you hiking? Lol, I live very close to ya. Looks like Umstead maybe 😃
When do you think biotech will take off? Probably not shortly after ASI
When do you think will we be able to reverse aging? I mean not only longevity escape velocity (that might be close) which is about lifespan, but in the sense of really reverse the damage in our biological to the point that it is visually as well as functionally negligible. So people essentially can turn back to their early twenties
what is your age?
Does anyone have suggested readings on 'the 4th industrial revolution' or 'post labor economics'? Need to put together my summer reading list.
Zeihan style
Hey David. I'm interested in your thoughts on AI at the "Edge". TCY broached the capability with a comment about QCOM.
Looks so peaceful out there
thank you for this.
What is your opinion on Metaquest os?
Do you expect the AGI this year to be recursively self improving?
Close. Next year definitely
Thanks so much Dave
@8:05 curious why you didn’t included Tesla? And why Boston Dynamics over other robotics manufacturers (including Tesla)?
Boston Dynamics has been doing robotics for like 2 decades +
Very good summary and refreshing because it isn't IMHO hyperbolic.
Love your videos love from Portugal
Its going to be amazing the future i cant wait for agi and good humanoide robots etc...
Please keep making videos like this I love hearing you talk about science I admire how much you know about these subjects I'm schizophrenic and I can't retain so much information but I always try..
It's difficult to argue against my father, he's very pessimistic, but now I'm going to start writing down the strengths to defeat the master at home...
Dave, while your enthusiasm is infectious, I'm curious if you might be overselling the near-term impact of AI technologies like NVIDIA's Foundation models and Project Stargate. Given the complexities of real-world application and regulatory environments, could your 2029 timeline for expansive AI integration and ASI be a bit too optimistic? It's an exciting vision but might require a more cautious forecast.
Add an AI interface to the web, and AI will surf at the speed of light.
When AI agents start talking with other AI agents... It's so close I can almost touch it.
Singularity 27
@@Hippida Sounds like we're on the precipice of some truly wild changes! The idea of AI agents conversing with each other is fascinating and a little bit mind-boggling. 🤔
Very interesting videos on this channel. Havn't watched many though). Have you covered a question on what country (or type of country) it's better be a citizen or resident of when AGI happens, when loosing all jobs happens and so on. Looks like it matters a lot. I think it's better be the citizen of a country which benefits from AI - where the most benefitting companies are. But maybe there are more complex causalty and correlations hear.
All I can say is I'm excited to see where all of this goes. I think you're right about robotics and AI being an "iPhone moment", but I feel like your timeline is a little optimistic. Who can really say though with how fast tech is advancing and accelerating
15 years ago I was telling my management class at Texas state university that the robotics revolution would make the industrial revolution look like child’s play. However, the end result should be an advancement of human creativity beyond anything we’ve ever experienced. Actually, Fuller “predicted” this back in the 60s.
Your time estimates fit with my eschatological estimates.
Stargate is exciting, but what about Dojo? It should be done earlier, what will it accomplish?
I love the nature talk and walk!
Could you do a video on CREATION of new data structures? Is there a process for creating better and better datasets in massive quantities? Will this be increasingly possible in the near future with GPT5 and the other best models? I dont use the term “synthetic” data since it seems a bit meaningless, as all data is synthetic and also real and its a spectrum. The point is if there are reliable processes for creating data of such quality and diversity that new models can improve faster, and there will be true data abundance.
I think a tricky thing about AGI is that many people feel AGI should be human like Intelligence. For example you can give a task like learn to operate this computer and the AI figures it out by itself, improving and learning like the human brain. But in that case that AI would blow right by us because it would keep learning and also you can always scale it. It would mean ASI.
Therefore AGI necessarily needs to have some limitations, things humans can do but it cannot. But as long as thats the case its not AGI for many. So I am not sure if there ever be a moment where we say, yes thats AGI.
We might go straight from nah not quite AGI to damn ASI
I've always thought of AGI as being a model that can learn in real time.
That is. And, at the moment, it doesn't exist.
Id like to see all your videos done in the woods! Bravo
Hi David, wouldn't you agree that it's more important to improve much more quantum computing so that this AGI could improve much faster and thus solve all the other problems earlier?
That's making huge assumptions about quantum computing
I think the convergence of spatial computing/mixed reality devices and AGI technology will bring us to a new consumer electronic paradigm post smartphones and laptops. An AI powered augment shared layer over our reality is what will bring us into the new era. Once something like the meta ray bans can be as powerful as the Vision Pro, that will be an iPhone moment and new properties can emerge. Idk if most ppl will have real physical robots as a personal assistant but surely a cute augmented AI assistant in your smart glasses will be commonplace.
I’ve just saw the profluentbio team with their opencrispr project which is freaking cool
Using LLM and AI to design gene editor
Maybe genic therapy price will start to deflate
How can we adapt for the 4th revolution ?
1.
2.
3.
Such a beautiful park. Where is it?
Outside
Outside 😂
@@DaveShap I don't recall that server.
I absolutely love that he almost always cuts himself off on his sign off.
I’m glad that there are people who remained worried, as it takes all types and every opinion matters when humanity is on this broad of a chasm. We’re all diving in at the same.
I can’t help but be excited. for every problem, an equally potent solution will be developed just as quickly. Bring it on, it’s gonna be an adventure.
What about the singularity? After ASI there may be no need for such large data centers...
Most importantly is the S&P500 a good investment factoring in all this? Or is it better to focus somewhere else?
Hopeful, optimistic, and excited.
"...switch hands..." I've been there 🤣👋
Was curious on if you ever think we will get some type of control over sleep, more specifically if we will require less or none at all
To do what? Brood over unemployment caused mayhem and WW3? :D
How do we create an economic system in time ?
I like this style of video nice hike in forest with you.
Very interesting perspective. What i think is that i need to help my kids understand and plan for their future’s. The job displacement will be a major upheaval whilst countries, economies and Govts work out how to deal with a population losing their jobs and income. This is a huge challenge for us all and i think a UBI will be essential otherwise we are going to get anarchy and war
Still prediction AGI in september David?
Great perspective and insights. What is your timeline estimation for BCI & AR/VR/FDVR?
5-10 years if I had to guess
Energy cost and data center cost will be the biggest limiting factors to the rollout of AGI, even if it's already here
Fantastic video. Uplifting and fast but realistic. AGI and ASI taking 3-4 years to become widely spread is fine, so long as health companies use it to cure diseases in the meantime (you said all diseases cured within 5-10 years of AGI, forget if you meant that for the lab or for actual consumers and patients)
Agreed pretty much and quantum coming this year too and hybrid ai.
When AI can teach us how to truly utilize Quantum computers.... Don't think that is far off
Microsoft Teams backgrounds have really improved.
If Microsoft has an “iPhone moment” Apple is in trouble they’ve been pretty quite on AI front, are they potentially too far behind?
I think we will eventually be seeing household robots coming from Toyota and Honda. Exciting times 😊.
David you should talk to Daniel Schmachtenberger.. would be a great conversation. Just a thought. Thanks for the value you provide
It only makes sense
Hi David, how do you think blockchain and crypto is going to be integrated into society. I don’t think you ever really touch on it, but you speak about AI and quantum computing so??
Hi, would like to see another video, of talking whilst going through the woods or something like that, but just tip for the future, maybe you can slow down a bit when walking because as in this video you were walking quite fast, there was quite a bit of breathing inbetween words and it made what you were saying at some points feel a bitt rushed. it naturally made it hard for me to grasp every word you said, but aside from that, very interesting video and i agree with quite a few predictions. 👍 Big fan also, Keep up the good work :D
Who doesn't want to listen to a breathless bald white guy traipsing through the woods??
@@DaveShap haha come onnn being a bald white guy doesn't make the listening / watching experience any better or worse😆(maybe the bald bit is better as i keep thinking im looking at picard which is funny)
What about solving the alignment problem, and making sure that ai has near human ethics and respect for life
Would the advent of ASI not also accelerate all other fields - for example, energy production. ASI would solve the energy deficit issue through the application of more advanced science and engineering than humans could conceive of? Seems the definition of the singularity needs to be reexamined.
I'm probably wrong, but I don't think Ray kurzwells prediction of 2030 included synthetic data. I would think that changes things.
Where were you hiking?