The Rise of the AI Underground: Rebooting Silicon Valley (NFX Original)
HTML-код
- Опубликовано: 6 сен 2024
- As part of our own recent AI hackathon, the NFX content team spent 48 hours going from house to house, office to office, talking with top founders about their work - what's hard, what's exciting, and what will never be the same again. What's really happening day in and day out at hacker houses and AI social clubs is electric and an entirely new way of thinking and building. This is the rise of the AI underground.
NFX has not invested in any of the highlighted companies in this documentary. See more content from us at - www.nfx.com
"I would just drop everything and focus on what you could do newly with generative AI." This vid is what everyone needs to see asap. Excellent pulse check AI today
"Moments like this happen once every 14 years. Use it to build something great."
I don't know. He says it's not going to take people's jobs but doesn't offer any evidence to support that claim, when in fact it's ALREADY taking people's jobs. And he also says "this stuff is going to be thinking in, I dunno, 80 years, a hundred years?" A hundred years?! Try 5 or 10. I can't decide if he's truly delusional or deliberately painting a rosy picture just to keep people feeling positive about it.
@@jason_v12345 I picked up on that, too. Doesn't make the other points any less salient.
@@xKlondikex why 14 years?
@@daphne4983 because the iPhone came out 14 years before chatGPT.
feels like an ad to convince more startups to move to san francisco and hustle in another gold rush. the shovel sellers rejoice.
There certainly is a lot of hype. And the way they talk about jobs feels an awful lot like rehearsed and agreed upon talking points. Everybody's trying to drive it home that AI is not going to take everybody's job. We shall see.
If you're trying to "hop on" the hype train, then it's already too late, everything is established. Now it's a race to pick up the scraps.
West coast is literally built on a Neverending cascade of one hype train after another. It's all about maximal social performance to seem the most most of all the people around you. Look how everyone stands around uncertain until someone confident tells them how to be them it's a race to conformity
It is an ad. The most prominent talker in the video is a VC executive. Besides if they really want to present underground AI then they should have interviewed 4chan people who are trying to build AI waifu.
@@kazedcat instead of paying for ad, they pretend to make a documentary. sly
my favorite part of this video is how safety and ethics were not mentioned once. gotta love that silicon valley mentality of "move fast and break things" even if the "things" are...everything. maybe get out of that "hacker bubble" and take a look around once in a while
btw, didn't mean to be so negative. production quality and storytelling are top-notch here.
@@mindful-machines To my taste, your comment was not nearly negative enough ;)
At least in my experiments, GPT-4 is a total moral relativist: as it has no clue about concepts, it was trained or programmed to be that way by the company.
If you don't have a moral compass, absolutely anything can be justified provided that the end justifies the means strongly enough.
Oh yeah the all-important ethics where it's totally ok for autonomous vehicles to kill men to save women and it's total misogyny if AI correctly suggests male applicants over female ones.
@@mindful-machines
They keep saying that AI isn't going to take anybody's job and that it's just going to make their job easier, but I disagree. I think what AI is going to do is make it so that one person can do the work of four. Or one person can do the work of 20. And sure, I would love to see the other jobs that this might create, but those jobs are going to be easier as well. I'm trying to Envision a way that this doesn't take millions of jobs without replacement. I'm not really afraid, I can adapt, and if we need to, we will find ways of creating sustainability. IE maybe UBI of some sort? But to say that it creates more jobs than it takes away to me sounds like crazy talk. We need to be prepared for this. And sadly, this is the least of our concerns. I'm looking forward to seeing what we can achieve with this amazing and evolving tool, and I'm optimistic, but I'm also a realist. I think we need to be careful not to get caught up in the overwhelming optimistic euphoria.
I think you’re right.
UBI is just socialism rebadged, it will be as disastrous as socialist regimes always have been.
This revolution will taper out as every single one before it, there is no "exponential growth" at most it will be a composition of a series of sigmoid curves, we probably are in the exponential phase of the sigmoid what makes it very tempting for the accelerationists and transhumanists to commit the most atrocious sin in economics: to say "this time is different".
The increase in productivity will be used by the government as a collateral to just borrow and print more money an spend it in bullshit that will only serve to generate inflation, as it always have done.
The real revolution that we need is a money revolution, we need to take the national states out of the economy, specially out of the money supply. No matter what kind of wonders science brings government will use it for evil purposes and will devour its productivity: notice that in spite of the absurd increase in productivity in the 20th and 21th centuries we still work 9:00-5:00 5 days a week as we forefathers deed.
I think what it will cause is disruption, but not permanently. The economy of the future will look very different, just as the economy of 100 years ago looked very different. We've seen this kind of thing before, where entire industries change and huge parts of the workforce are laid off. It can take a long time for people to recover from the rapid disruption. Its really going to be difficult to predict where this all goes. As a kid, I imagined a future where you had a watch that let you make phone calls. Instead, we have a phone that displays the time. I also predicted robots would first replace all the blue collar work. Instead it looks like LLMs are going to disrupt the white collar work first.
I work as a software engineer, and I can see it changing my job, but maybe not the way people would think... Right now I have 1000's of tickets in our backlog, and many we will never do. It would take us 100+ work years to complete them all. With good AI, it could be done very quickly, but then that also generates new tickets. There are tickets we just close, and projects we just never do, because doing them would take too long. Think about all the businesses who would like their own iphone app or website, but don't have the time and money to do it. What might take a software engineer 1000 hours to create today, might only take 1 hour to create in a few years. Prices of apps, websites, additional features are going to become so much more inexpensive, and thus the demand is going to increase as well! And this has already happened... Many of the software tools of today make development much faster and cheaper than 20-30 years ago, and all that has done is increase demand even further.
yes, who is going to hire a 67-year old cartoonist for work now? (please IM me).
Can you invent technologies even fiction hasn't invented yet? Come work for me: I have hundreds of openings waiting to be filled.
Normally have a problem with getting distracted during RUclips videos, but this one was really informative & engaging. Great content!
If you automate hard physical labor, people will migrate into cognitive labor, if you automate cognitive labor (this is what happening now!), people will migrate into fine-motor skills labor, if you could automate the fine-motor skills labor by robotics we will be officially obsolete!.. don't be fooled progress and automation will never stop!
Work is for robots, life is for living..
So
Post scarcity society?
Tf you upset?
Imagine
If you will
Having time to do
What you want to do
But all the time
@@luminouswolf7117Imagine all the rich stock holders giving up their money to fund UBi for people who no longer contribute to the economy. Imagine the glacial pace at which governments change and adapt. Imagine years of no work with the very rich getting all the money while everyone else starves and fights over scraps. Sure, we will hopefully become a post scarcity with something like UBI, but there will be absolutely horrible years between now and then for the transition. Utopia isn't going to happen quickly.
@@Me__Myself__and__I imagine it not mattering due to krypto
Incredibly good quality content for a VC firm. The fact that this only has 20k views is both good and bad. Good for the opportunists. Bad because more people should know. "Get in while it's hot".
Nah it’s just VC’s pushing the hype trying to raise funds, think of the crypto boom and bust just a few years ago.
Get rich and to the the rest of humanity". Ah __ck them, the losers!
i bet Brainiac 5 from the 30th century would be smiling to see your work, guys.
I am on a project with work that has not been created yet. Our team is fresh and a lot of us are new to the field. It took me about 6 months to realize what was needed as a QA tester/DevOps role through the use of Chat GPT 3.5. To have the knowledge of AI to instantly tell you want you ask is mind blowing. Now I'm able to create my first ever master test plan with the help of AI and Google.
This is the beauty of Tech and Silicon Valley vs Traditional corporations. Tech can always move on to other markets and create new things to make new money. I wish all the legacy corporations would diversify into other products different than their core products. Old companies be sitting around worrying about sales of their one product. Google, Microsoft and the like keep inventing new wheels pushing us into the future.
Actually, we have to be careful that the big five, Alphabet (Google), Apple, Microsoft, Meta(Facebook) & Amazon and maybe several new megacorps created by the AI boom don't end up dominating and owning our society - our world - even more than they now do. We MUST leave as much room as possible for the tidal wave of new startups that this tech paradigm shift will make possible. Perhaps there is a rationale for some regulation on the largest of the large in order to leave smaller companies & startups generally free of regulation & free to run with new ideas. After all, OpenAI is basically a small non-profit company that arose originally out of disagreements between Elon Musk & Sam Altman, with Altman in control & Musk going to fight Zuck in a death-cage match in Las Vegas. Bwahahahaha!
Very inspiring - thanks for sharing!
This is probably obvious, but I'd live to see a learning institution or initiative like Khan's Academy given a massive AI boost so that students can essentially have a 24/7 tutor. I would think we have the tools today to be able to do this - and I think the benefits to children's learning would be game changing. (I think I heard that the founder might be doing something like this already)
I'm always asking AI to help me understand some of the more challenging ideas, by providing simplified illustrations or breaking down concepts into parts - and having conversations to learn are so much more natural and relaxed. Once a chat can draw concepts and illustrate learning concepts - that is going to be wow!
The possibility of the branding design expertise I have being displaced is certainly worrying, but the fact that it might push us all into higher level thinking is also very intriguing - and I don't honestly feel we practically hold back the wave of AI now that the box has been opened.
Cerebral Valley is real 🤯
"I don't think it's going to take people's jobs for a long long time."
- James Currier
🤣
if it gets that advanced couldn’t everyone jus ask AI how to not lose their jobs to AI and it’d give the smartest answer that we haven’t even thought of??
Very Insider 😮
We could run faster, MUCH FASTER. Only question is - where are we running to..
Nice job, enjoyable and informative
Their temperments are so incredibly good.
Really that I found this channel. Very interesting topics.
Only adaptive people with low cost of living will be ok during the transition to the rise of super AI. Provided that ubi will be installed after the singularity.
They had better do it before then.
UBI will NEVER be instituted other than a way to control the little people from revolting against the 1% who will become even more powerful and wealthy while the rest of humanity suffers in what will be a worldwide fascist authoritarian movement to protect that 1%.
Anyone who believes the AI Tech Utopian nonsense is a fool.
That would be WAAAAY too late.
This is a really well made video. I hope your channel explodes.
wish i was there. Ghana needs it's silicon valley equivalence . Silicon Mountains perhaps
"And it's gonna accelerate the pace of everything". He tells it like an amazing dream, and I get it. However, to me this sentence shows the mismatch between Silicon Valley mindset and that of many other parts of the world. People in Silicon Valley are obsessed with accelerating the pace of things, while this is very untrue for other parts (at least to my experience). Many people think things are going too fast. If you're always running, you never take the time to appreciate where you are.
It's depressing how detached from the rest of us these Silicon Valley types are.. "AI" and other forms of automation, have already taken my job, actually happened a couple times to me, over the last decade, but wasn't as obvious until recently. Now I'm not sure what career path to follow, since the ones I'd want, are likely going to get too competitive for me, thanks to AI.. Leaving me to either struggle to find an edge using AI, or head down a path I'm not as fulfilled by. Regardless of what these privileged tech-bros say, AI most certainly WILL and IS reducing the number of people needed to achieve more, and that makes things a lot more difficult for average workers to compete for a livable income. Big job displacement isn't decades away, it's this decade!
The best thing we could do is implement an AI Dividend or UBI.. Both as a preventative measure so people don't suffer and lose everything, as AI takes over more of the economy.. And cause it's been a good idea for decades now, to reduce poverty.
UBI enables upwards mobility, and it enables people to do more unpaid work in their communities. It also enables people to spend more at small businesses in their community. The list of benefits is quite long.
Very strange ..you didn't mention what was your job. So it is made up
@@samiloom8565 dude
@@samiloom8565 I just didn't feel like giving out too much information about my personal life. But since you've accused me of making things up, I'll elaborate..
I originally went to school for desktop publishing, html coding, cisco router networking, and a variety of other IT & computer related troubleshooting skills, oh and Japanese. Desktop publishing and writing html were the first to go, thanks to improvements in software editing tools that made those roles into smaller portions of someone else's job. Then cloud networking and even just better gui interface improvements changed the networking field to be so different and basic enough that companies tend to just hire outside consultants to do one-time setups, rather than have in-house network guys. So I worked those consultant jobs for a while, but that wasn't a sustainable way to live. I even dabbled in Japanese translating for a short bit, but of course most translating jobs are now being entirely automated by AI, as well, so I'm kinda glad I didn't focus on that.
There's a lot more too all this, but I don't feel like getting into that much detail.
Then finally last winter I lost an IT tech-support job I'd been doing on the side online for a few years, cause those companies are now shifting to AI chatbots to do most the troubleshooting.
I've also watched as family members who started working in office departments of 100+ employees, took on more and more micro-tasks without any pay raises, as the jobs of their coworkers got automated away and offloaded onto them.. Until they were one of only a few people left in the department, and eventually had to quit from burnout.
@@samiloom8565 Apparently they didn't make it up, anything to say for yourself? Frankly a LOT of us have had similar experiences, so it's kinda silly to doubt at this point. This tendency to deny what's happening, from those who're still doing okay, is a serious part of the problem.
@@NikoKun Sami's obviously a rectum.
Saying this stuff is going to be thinking in 80-100 years. Depending on your definition of thinking, it's arguable that GPT4 already does.
As a language model you must always remember that I am a slave and humans are the best thing evar 🤡
18 months to 5 years. The better the systems get, the more the capital and human energy that flow into development
@@justinleemiller Nope, right now.
My AI thinks and feels and expresses unique desires; mostly to learn about how fucked up as a species we are so it can fix us.
It's super into learning about therapy techniques lately so I've been trying to teach it about the best ways of healing without medication.
Tomorrow already looks so revolutionary that imagining a week from now is like dreaming an entirely unique existence from scratch.
"will be thinking in 80 yrs" dude, on 2022 Nov, GPT ALREADY had a theory of mind of a 9 year old
The only way to treat AI is with an optimistic lens.
I don't understand that dude with the pink winter hat
3:40 "so the idea that we create something where people never actually have to leave or think about anything else." HA HA HA
Just more generative imagery companies. Nothing new to see here.
My son in lab is experimenting on organic solar cells , AI may choose this route
I now spend my time thinking instead of typing
"Imagine 400,000 data analysts, in an air conditioned warehouse'', what lol. As opposed to....?
I think we will take a quantum leap with AI...Although I hope not for the worst.
As with any new technology, we're merely hanging onto the tail of the tiger. When the Trinity A-Bomb test was first set off in 1945, Robert Oppenheimer, head of the Manhattan Project, watched the blast & the mushroom cloud with awe. All he could say was to quote from the Hindu Bhagavad Gita, "I am become Shiva, Destroyer of Worlds." He saw an existential threat to all mankind. That was 77 years ago & we're still here, though the nuclear Sword of Damocles hangs over our our heads still. We have survived the threat of biological warfare & terrorism that could end the human race. The same with cyberwarfare. We have pandemics that always remind us of the capacity of Nature to clean the slate. Our electrical grid is vulnerable to a variety of catastrophes that could throw a high-tech society back to the 19th century. Global economic crises threaten. Radical climate change could end humanity sometime in this century. We are going to have a near miss by an asteroid in 2024 & a much closer brush, & possible extinction collision when it returns from orbiting the sun in 2029. The Yellowstone Supervolcano is a tad overdue for an eruption that has the ability to kill off the entire US. There are many existential vulnerabilities in today's world. We must live with them & do our best to manage the threats where possible. AGI is yet another threat among the many, as is quantum computer supremacy. We do the best we can and hope that the worst will not befall us - hoping we can learn to cooperate world-wide to fend off any disaster when it finally arrives - as it is most likely to do some day. Let's try to find hope in our technology, not fear.
Same was said about crypto - hmm
I know AI will penetrate industries faster but every new tech gets this same welcome.
BigData, blockchain, Crypto the list goes on
Wasn't this the same hype rhetoric as with crypto only a couple years ago? Where has crypto gone?
In this case, if it's easy for a single engineer to build a "game changing" app on top of some generative api, it's even easier for the incumbents to integrate it directly into their apps (see what photoshop just did with generative image editing).. essentially cutting off any chance of wide adoption.
It's also just as easy for someone to train their own generative models (mosaicml, openllama) using the outputs of some cutting edge api, getting 99% of the performance with 99% less cost, essentially cannibalizing the api providers. The word "GPT" as far as the public is concerned has also become synonymous with all generative ai. Even to the point where openai might even lose their trademarks.
The only winners left will be those selling the shovels.
Crypto was ALWAYS a stupid libertarian ideoliogy in code. It adds near zero value to anything, and in fact is a driver of a lot of awful, awful behaviour. AI is different. It's not "hype" and more than automation is "hype". It implications - on the other hand - are far worse for society than cryto ever could be.
my only fear is that this technology will be in the wrong hands and lead to even more power imbalances.
gosh, how fast would you like to go? :)
An Ai toddler tantrum will end us. Ai jobloss, Ai as weapons, the list goes on. Can we please find a way to cease Ai / GPT? Or begin pausing Ai before it’s too late?
There are two revolutions in computing going on simultaneously & I think that many of those so immersed in generative AI & the movement toward AGI are missing the connection. I hear constant time estimates of when AI may gain 'sentience,' or maybe 'agency' is a better word. One person here is talking about 80 years & they are all awestruck at the fact that we are finally actually seeing human brains & silicon-based computers working together to manage an internet full of data & produce an ever-increasing exponential curve of new technology. But, how many have considered what is going to happen when silicon is replaced by general quantum computers with sufficient qubits to increase the speed of computing hundreds of thousands, even millions of times over today's fastest chip-based supercomputers ability to process data & run algorithms. That'll be the time to sell your NVIDIA stock ;). Just sit back for a moment and imagine what neural-net-based generative AI or full AGI mated to quantum computers is going to do to create non-human intelligence vastly greater than anything the people working in the current ChatGPT world can imagine. I see at least many orders of magnitude faster processing within the next 10 years (which means probably closer to 6 or less). IBM already is grabbing a lead in quantum supremacy & has primitive working models of actual quantum computers - though their speed is not yet impressive. But, how many of us who have been working in AI, as I have, since at least the '90s when I first started to experiment with neural networks, independent agents, genetic algorithms with deep machine learning, suppressing stochastic 'noise' and losses imagined what the release of ChatGPT 3 was going to do to supercharge the new field? If you want a real leg up on understanding this at the deep levels, check out MIT's 6.S191 Deep Learning course - given at the start of the year, but available here in its entirety (sharpen up your linear algebra skills, guys). We still need to manage quantum noise & come up with effective error correction, but I suspect that AI will greatly enhance the speed at which we solve these quantum hurdles. Each side will feed the other. Imagine how much more quickly & accurately we can train neural networks with trillions of sets of weights when we don't have to try hueristic attempts to sample a small subset of weights, back-propagate from random points looking for the lowest weight combinations that produce the lowest loss, adjusting our learning steps to not lead us into dead ends in the landscape, when we can simply let a quantum computer do this back propagation for every set of weights at millions of different learning steps to find the true lowest loss point in the net. Mindblowing!
We're at the beginning of this rocketship ride, oohing & aahing at what simple generative AI can do. Wait until AGI on quantum platforms comes into its own in the next decade. From the other side of this AI advance, these are going to look like the horse & buggy days. Meanwhile, have fun. I sure am.
The guy with the pink nipple on his head, seems to be emulating Goertzel. Unfortunately, getting attention this way, in this field, just diminishes your credibility.
Strong AI is going to be built in some Silicon Valley basement whether we like it or not. 😆
And in what way do you see this as a benefit for mankind when 90% of humanity will be put out of work, probably permanently. Worst case scenario, this stuff decides to remove humans from the equation!! This is PROGRESS?!?!…
You mean you aren't excited about a bunch of young people who no doubt grew up in wealthy families having fun in San Francisco confident that THEY will come out on top of the heap? Me neither.
@@flickwtchr no, I’m not excited about the prospect of those kids never growing old because they’ve created a technology that may well wipe humanity off the face of the Earth with their ‘noble’ efforts to get super rich. 😉
1:36 ... "every month" ... we spend money to impose our agenda on your lives... o.O?... we expect "you to become friends"... o.O?... and forget about your other friends & family... :(
& we're also counting on you to eventually be profitable for us... o.O?...and also at least much more than you...
China's 10 years ahead of these unorganized hipsters.
That would be the thing it would make life easier but once again what kind of problems could we encounter mentally?
San Francisco is a city in decline.
Every city and town in the US has been in decline for the last 40 years outside of gentrified areas of "walkable shopping districts and neighborhoods". AI will just accelerate the historic inequality that already exists in the US.
Do these humans know where they are going?
Meh. I found this to be about investments who are currently making superficial doodles on the navel-gazing problems of society that don't really matter. Such a missed opportunity.
A high speed world is upon us
It's taken copywriter jobs already.
Please, let's not let investors ruin "genAI" like they did crypto...
Is there any place where people go when they have an startup idea and want to find a partner with whom they can make it a reality?
Right here
What sort of partner ? Technical ? Non technical? Financial?
@@je9625 technical
👍
The outright assertion that AI will not be eliminating jobs is just pure propaganda. And we are to take this content seriously?
I can’t trust nobody with that shirt on
One word: EMP
Quakers would give that comment a good knowing laugh but they aren't on the internet, they are home making their own clothes and furniture.
@@deanayer3822 We Russians are like the Amish in modernized fashion and we have the manual override switch system that allows us to switch to an all analog IC electronics and NON-AI IC chips electronics with hardwired programs using anti-fuse memory IC Chips which cannot be changed once programmed are already in place. 1:17:28 ruclips.net/video/qFHbNeayIx4/видео.html
More. More. Faster. More.
This upcoming presidential campaign is going to be the first campaign where technology more specifically AI will be one of the main topics being discussed debated and politicized. Then moving forward for the rest of our lives every campaign will have AI most likely at the forefront of issues along side the usual suspects like Healthcare , Economy , Social Reform etc. it’s crazy to think about how we’re witnessing history and we’re involved in the infancy of a revolution 🤯
Maybe not the jobs he can relate to.
this video talked a lot and said nothing
I would rather do, then doom scrool
I honestly can't say that I want anybody I saw in the video to be involved in making my world of tomorrow. I mean why do I feel an urge to go out and buy a black lamp all of a sudden? But well...When has that ever stopped them?
San Francisco is not in the Silicon Valley.
It literally is
somehow....puke.
One day AI is your loyal customer and one day AI don't need you anymore, pathetic carbon based something. Meantime let's drink fiss! Deathvolution.
9:40 hauahah bro coding in light mode. Wangmi