Jaron Lanier Looks into AI's Future | AI IRL
Вставка
- Опубліковано 17 тра 2024
- Jaron Lanier is among the original Silicon Valley utopians. This ultimate insider gives an unfiltered look into what lessons tech booms past can provide us about artificial intelligence's future.
--------
Like this video? Subscribe: ua-cam.com/users/Bloomberg?sub_...
Become a Quicktake Member for exclusive perks: ua-cam.com/users/bloombergjoin
Bloomberg Originals offers bold takes for curious minds on today’s biggest topics. Hosted by experts covering stories you haven’t seen and viewpoints you haven’t heard, you’ll discover cinematic, data-led shows that investigate the intersection of business and culture. Exploring every angle of climate change, technology, finance, sports and beyond, Bloomberg Originals is business as you’ve never seen it.
Subscribe for business news, but not as you've known it: exclusive interviews, fascinating profiles, data-driven analysis, and the latest in tech innovation from around the world.
Visit our partner channel Bloomberg Quicktake for global news and insight in an instant.
"One needs to be a critic in order to be truly hopeful." A great insight.
Misquote
He is one of those people who not enough people have heard of, but who will be remembered as one of the greats. Well done for having him.
Right. Overappreciated.
I remember reading an article on him in Wired forever ago. Was on my mind of late. Glad to see him still part of the conversation. He makes some truly salient points.
I'm so impressed with the hosts! Both are so smart, thinking and responding on the fly, and engaged. Great interview
I was thinking the same. Especially given how vast the topic is and how much Lanier is covering, they did an excellent job of grabbing certain moments for further clarity or followups.
Well, in fact it is on some independent podcasts where you could find the best hosts for interviewing the smartest people, and within a much longer format.
The hosts are great actors. Lanier does well, but still gives off the vibe of someone trying to explain something to Elmo's less intelligent cousins.
Is this satire? I'm 3 mins in and the host has already interrupted him - 3:27
The hosts really pressed him on some important points. That was very impressive
a note about the CC subtitles: what is written is "Providence," but Jaron Lanier is actually saying "Provenance" - in the framework of Data Provenance. not sure if it was pointed out in comments, and just in case someone needs clarity there, because this is probably the most important point Jaron Lanier is making.
Jaron Lanier is an underrated voice (probably because his message would cost Social Media companies a lot of money!)
They'll cost themselves a lot of money if they're not careful.
He is indeed an underrated voice, but the conversation was somehow pretty short and I would love to have watched more
What, you mean the guy who wrote a book called “Ten Reasons to Delete Your Social Media Accounts Right Now” might be less than ideal for the interests of social media companies? 😂 (Great book btw!)
The reason why you do not know him from the media is his looks. How many people like him do you usually see in the media? It's our superficiality. We want to see nice, lean people in a suit or sports outfit coming right from the hair dresser, so that's what we (usually) get to see.
Google has 1M results for "Jaron Lanier" and 64M for "Bill Gates". If you take into account what I just wrote then 1:64 ratio is not that bad… if you accept a google search result count as a measurement.
Someone can only be valued if what (s)he says is known to people. Now he and what he said/wrrote is not known that much becuase .
I've followed Jaron since the mid 80's because of his ability to make the extremely complex easy to understand, and his stance on technology oversight for transparency and fairness. I think if tech had more people like him, we'd likely be a lot further advanced than we already are with considerably more useful technology benefiting society, which has always been his aim.
I agree.
You mean business had more people like him? Tech people generally just do what they're paid for.
@@ericcricket4877 Both Businesses and tech people. If there were more tech people like him, there would be more businesses that think like him (IMO).
I think his argument about open source falls completely flat simply because closed source has lead to the very thing he suggests open source would do. I'd much rather live in a world where everything is hyper centralized and open source than hyper centralized and closed source.
It’s underrated just how much he can dumb down these topics for us regular people. I’m sure when geniuses talk to each other in private, it sounds like another language.
Bring this man here again, and each time will be unique and full of freshness.
Well done Jaron!
he's a rare genius and I have followed him long enough that his speeches turn out to be extremely prophetic. He's like a hippy version of Ray Kurzweil.
The way he smoothly glided into declaring that open source = monopoly and then using Google as an example of open source demonstrated that he is not working for Google but a competitor, and no, that observation is neither correct nor accurate in any way. It was a slick try but it didn't pass muster. Google itself is not open source GNU/ Linux, Firefox, Gimp , VLC etc are open source. He avoided the real examples
With his pedigree that was on purpose. So be wary...
He is a snake and should be treated as such.
Yes, he works for Microsoft@@hitmusicworldwide
Why is this guy so likeable? The way he speaks, the explanations, they're so engaging, made me want to hear more and learn about such a new topic like AI. What an amazing session.
Because he smiles and speaks in a friendly tone. Typical human easy to manipulate
cause he's a musician
you learn how to never sound boring practicing music, its a very very useful 'hobby', it levels up every aspect of your life
If you like listening to him, I highly recommend his book, You Are Not A Robot.
Because he knows the cameras are rolling.
Yes! Just a slight correction; the book title is 'You Are Not a Gadget: A Manifesto' @@chrisklein9618
When I lived in the Bay Area in the 80s I remember Jaron being a big deal in the hobbyist computer world when he was really pushing the boundaries of the limited tech that was available then. I really appreciate that he’s still at it today and is great at taking that 30,000 ft view of the coming revolution.
The interesting thing is that Microsoft hired him to be an independent dissenting voice (so he told me). Much like Minsky, they actively want to engage in the argument. That is encouraging to me.
I love Jaron every time I hear him talk and always end up feeling enriched
Same! His views are always unique and very well defined.
I think he counts on math so much and I intend to do the same in my studies.
Jaron Lanier is a legend; he has some great books and I cannot recommended them enough as he's just such a thoughtful, insightful dude. A particular favourite of mine is "Ten Arguments for Deleting Your Social Media Accounts Right Now". The title, you would think, says it all but that book is so full of amazingly erudite observations and suggestions. It changed my life in the sense that it framed for me the acerbic toxicity of social media so succinctly that I never looked back on abandoning every single shred of social media interaction in my life.
nice! to learn there a book like this. i’ve ditched the digital socialization (other than YT i suppose) and haven’t missed it much!
I’ve read that book and subsequently deleted all my social media and I can say that that has been one of the best decisions I’ve done for myself and my well-being in a long time.
I can actually see clearly, looking back and in the present, how social media has had such a negative impact on my life in many ways.
I can actually look back and see how manipulated I was, how addicted I was to those platforms, how much it controlled my life and my thoughts, it’s scary to even think about now!
Anyway, anytime I see someone who’s read Jaron’s book or talk about how they have deleted their social media it makes me more hopeful.
I'm a little confused by what you said.............this isn't meant as sarcasm or to be disrespectful but isn't UA-cam a form of social media, and if not why don't you think so?
I have social media, but I don't really look at it. I have one real friend.
But I like things like discord cause talking to strangers, international is something special to me.
Things like tiktok or insta on other hand are a mirror of the society and an indicator for how important real education is.
I was skeptical of Jaron, but he has mostly sold me that he is one of the few people who understand the real issues.
why were you skeptical? Are you just now learning who he is? Dude has been one of the most truly knowledgeable tech people for 30 years
@@humanbeing5300 Because plenty of knowledgeable techies are either corrupt, confused, or most often both when it comes to the real issues in ethics.
@@konberner170 we need more Jaron’s and people like him that’s for sure
@@humanbeing5300 In my view, the fact Microsoft has him is a solid +1 for them.
Holy cow! This guys looks like he would be speaking well above your head, but man, he speaks plainly and in a way that a regular person can understand. He’s got a very great skill in this.
The moment you notice jaron lanier has some new content 🥳🥳
Awesome conversation ! We want more Jaron !
Its so nice to hear brilliant ppl like him.
Interesting interview but an improvement for next time would be to have at least one interviewer who understands on a technical level how different models work. Recommender systems are not at all the same thing as large language models, like saying cardiology and pediatrics are the same thing because they are both "medicine".
I remember him from my Mondo2000 days, true pioneer who would have steered computing in the right direction if it had not been high jacked by capitalist
I have faith he’s pushing envelope from the inside.
I was at the screening of Minority Report with Jaron. And had set his virtual reality headset on my head ten years before that. And, like you, I was in the Mondo 2000 bubble.
Jaron was correct on every point in this interview. And the interviewers were on the ball, and actually listened.
Thanks for Jaron and the 2 hosts for this interview, it gives me hope to see humans working on the future ways forward
People who know there is no black and white are great. I like these kinds of interviews. Regulation is very important.
Jaron has been a pioneering voice on tech for decades.
Great video! Please more from Jaron Lanier!
Yes, a very nice conversation, I especially liked the idea of a strong legislative structure being able to actually facilitate innovation, having never thought of it that way. The other lovely part was about humans liking to make things which others wil buy, because in order for one to buy, one has to be willing to give up a part of one's living (as in 'making a living').
Every time i hear Jaron speak I wonder how he managed to roll nat 20s on Wisdom, Intelligence AND charisma. I think hes a benevolent matrix agent insert.
Criticism is the true face of optimism, and courage is the true face of criticism. Live with a backbone.
This was a great conversation, would have loved to see Jaron on some longer form podcasts. This shouod be watched by more.
Most insightful talk I heard for a while.
Thanks for having him on! A true pioneer and voice of reason
Great interview, thx for making it👍
This is the most coherent I've ever seen Jason. Great to see.
"The Critic is the True Optimist" - this defines my entire career (and my success in it). Being brave enough to point out (and then fix) problems is critical for tech.
Great interview.
I wish they had a more refined discussion on the question of open source. the interviewers didn't seem to fully understand the arguments for and against it and didn't ask him to clarify his stance in a meaningful way. eg, the question of what's going on in the "black box" of AI is not the argument for open sourcing it. it's about control over and knowledge of the training data and output. with fully proprietary closed AI models, censorship becomes a much bigger problem.
Wow... what a great talk. Like I'm glad I watched this.
Great conversation. Thank you for sharing.
Very interesting and thought provoking, thanks!
Well done ,,,needs to be heard .
Thanks♾️
Wow, a MSM interview where the interviewers actually did some research!! Very nice.
I can smell that dude's head through my screen.
Brillant Interviewers - brilliant interviewee ❤👏🏻
Wow. Did not know of him. It has been a long time since I have heard thoughtful, knowledgeable answers.
I love that Jaron is still involved in a meaningful way in industry. It would be a shame if he were to throw up his hands in despair and leave entirely due to frustration with [insert any typical reason here that creative types leave big companies]. So much better to see someone like this trying to guide things from inside.
Why did you cut in the middle of him speaking? What did he say that you left out of this version of the video? Just curious.
Can somebody explain to me his argument that open source leads to hyper-centralization. He said its due to mathematics and that there are numerous examples but he didn't explain the math or mention an example.
"Releasing the code supports emergent monopolies"
That struck me as weird too- but ultimately more revealing about his worldview than anything.
I’m guessing he means that psychopaths like the ones at Microsoft will weaponize people’s altruism against them by acting in a zero-sum fashion. Probably because he works at Microsoft and knows the mindset well, lol.
This interview was scarier
Had me confused as well. So I asked ChatGPT & its answer makes sense to me.
It's up to you to dig deeper and weigh its validity, but it's a starting point to explain what you asked. Here's what it wrote:
Jaron Lanier's argument about open source leading to hyper-centralization is a complex and somewhat controversial perspective. The basic idea is that while open source is intended to democratize software development by allowing everyone access to the code, in practice, it can sometimes lead to a few dominant entities controlling a technology or market.
Here's a simplified breakdown of the argument:
1. Open Access to Code:
Open source software (OSS) is freely available for anyone to use, modify, and distribute. This theoretically allows for a wide range of contributors and innovators.
2. Skill and Resource Disparity:
In practice, though, contributing to and effectively using OSS requires significant skill, resources, and time. This means that while everyone has access to the code, not everyone is equally equipped to use it or contribute to it effectively.
3. Dominance by Few Entities:
As a result, entities (like large tech companies) with abundant resources and skilled personnel can dominate the development and direction of key OSS projects. They are better equipped to contribute significant improvements, maintain the software, and build infrastructure around it.
4. Dependency and Control:
This dominance can lead to a situation where a majority of the market becomes dependent on a few entities. They essentially control the technology despite its open-source nature, leading to a form of hyper-centralization.
5. Mathematical Basis:
The mathematical aspect Lanier refers to might be related to network effects and power law distributions, common in digital and networked economies. In such systems, small advantages can compound over time, leading to a situation where a few players hold a disproportionate amount of power or market share.
While this perspective is subject to debate, it highlights an interesting paradox in the digital economy: initiatives aimed at decentralization and democratization can sometimes inadvertently lead to increased centralization.
Btw, OpenAI started as a small open source out fit, but the road it's taking, it is showing signs of an emergent monopoly. It is well on its way to becoming Big Tech.
@@kamu747He seems to be mingling distribution of source-code with distribution of power (I think those two should be though of as separate). In linux power is centralized on Linux Tovalts. Although his project is open source he retains control over it (I think the community yields to him out of respect). Lanier's argument doesn't fully hold. There are many Linux distros. There are many browsers based on Mozilla code. OpenAI is showing signs of emergent monopoly because it abandoned its open source nature to retain power. I think when Linus Tovalts passes away I say the tower of babel syndrome will play out (due to a power vacuum) and Linux will start to fragment.
@@kamu747Its like the failure of communism. Wealth and power as two separate forces. Communist believe in common prosperity by supposedly allowing everyone to share the resources, sharing the wealth. But what communism governments end up doing is horde power. Communist countries have a large power gap between the haves and the have-nots. hence the imbalance, hence their problems. Capitalism is sorta different. Capitalism promotes hording wealth but democratic governments tend to better distribute power with the people. Share the power. Its better to have power then wealth.
I love this guy. Jaron is very impressive.
I like how he says "we" when taking about big tech. It helps remind me of potential bias but also the viewpoint he is coming from.
Bloomberg did amazing with this interview
What a gold nugget this guy is! Reminds me of an old friend of mine.. Anyways, from the get go "Criticism is the true face of optimism" that centence alone has more wisdom in it than like anything I've heard or read this year so far. Thank you Octopusman!
What a wonderful, knowledgeable and pleasant person. Loved the interview
Excellent talk! Tanks so much!
Thanks for the open and honest expressions.
I thought I wasn't going to like this guy, that he was not what he's being paid to be. But then at the 12 minute mark I realized this guy's a genius! Geniuses really do have weird personalities - it's just a fact.
Immortality of monism through the the perspective of computing experts.
Jaron's notion that human generated data will continue to be relevant to AI training as AI models progress, presumes that it is not possible to transition to AI self training in some way similar to the step between AlphaGo and Alpha*. That may turn out to be true for some time to come, but the mere existence of organic intelligence suggests that those days are numbered.
Once a general AI architecture no longer relies on humans to generate training data, the shift of human workers toward a service economy (even including things like Etsy) will overtake all of us. At that point the only way to keep pace with AI super-intellect will be to hybridize yourself so that you are no longer bound to human intellectual limitations. I think the only way to avert massive levels of poverty and desperation - without essentially eliminating humanity - will be to offer a generous UBI, paid for by the hopefully dramatic increase in productivity that automation brings.
This is NOT just another Industrial Revolution. A hundred years ago, vast numbers of blacksmiths retrained as auto mechanics, but that revolution was not about people. It was about physical work. Horsepower. And virtually all the horses very quickly lost their jobs.
It's going to be a wild, scary ride. And we're already boarding.
Most of the top-of-the-leaderboard smaller language models are trained using synthetic data created by GPT-4 and have seen massive performance improvements. It’s already happening.
The conversation was spread too thin I think. Some depth on specific topics is what I came for.
I've been following Jaron just to hear his thoughts on social media and AI. He was always spot on and provides great insight into the problems with them.
Jaron Lanier is superchill! And the hosts did their homework well.
Jaron Lanier's contributions and perspectives are particularly relevant in the current scenario for several reasons:
1. Ethical AI Development and Use: In an era where AI is becoming increasingly integrated into various aspects of our lives, Lanier's insights remind us of the importance of ethical considerations in AI development. His emphasis on data dignity, where individuals are recognized and compensated for their contributions to AI training, is crucial for ensuring fair and ethical use of personal data.
2. Critical Evaluation of Technological Advancements: Lanier's role as a critic, despite being a pioneer in the field, is essential. In a time when technological advancements are rapidly evolving, having a voice that not only celebrates these developments but also critically evaluates their impact on society is invaluable. His perspectives help in balancing the excitement around AI with a necessary caution about its potential negative impacts.
3.Advocacy for Transparency and User Empowerment: Lanier advocates for transparency in AI systems and empowering users by making them aware of how their data is used. This is particularly important in an age where issues of privacy and data misuse are prevalent. His views encourage a shift towards more user-centric models in technology, promoting greater trust and accountability.
4. The Future of Work and Society: Lanier’s ideas about compensating individuals for their data contributions tie into larger discussions about the future of work and the economy in the AI age. As automation and AI continue to transform the job market, his proposals offer alternative ways to think about economic participation and the value of human contributions in a digital world.
If we do not heed Lanier's insights and warnings, we risk developing and deploying AI technologies without fully considering their ethical implications, societal impacts, and the need for equitable data practices. This could lead to increased issues of privacy infringement, unequal power dynamics in the digital economy, and the perpetuation of systems that prioritize corporate benefits over individual rights and societal welfare. Lanier’s contributions offer a path towards more responsible and beneficial AI development, emphasizing the need for a balance between technological innovation and human-centered values.
This was written by Chat GPT?
@@realfarmerdelatorre sure reads like it
@@SR-fx5sm It does, and yet i find it to be a valuably pertinent reflection / contribution (was glad I read it).
wow, blown away by his intelligence and great questions from the hosts picking up on things said and extrapolating on that.
I'm trying to understand this idea of open-source, hyper-centralization, and the exchange of information being the monopoly. I think what he is saying is that without monitization and market forces, there is no competition. That's what creates the monopoly. Really that's brilliant.
Excellent discussions
'Either the customer is you, or it's someone who wants to manipulate you.' - YES x 1000
His book 'Dawn of the New Everything' is incredible and prescient (published in 2017). He is one who eschews social media but I try to share his work in VR and thoughts on AI and tech whenever I can.
This is a hilarious interview. They ask him a question and he just goes off in a completely different direction (over and over again).
He seems to be there to try to justify the closed source nature of Microsoft
Thank you for your insights, hostile Chatbot!
Not really. He seems to answer the questions.
It's called being evasive.
I agree. He seems actually not so clear making his points.
If only the hosts would stop interrupting his answers to their questions, we could learn so much more.
"We don't want to mess up society. We depend on society for business."
Great Video/ Interview. Jaron is a wise man! 👌🏻
Love listening to him speak. Fantastic.
The privilege that you can create a position for yourself in a company you work and name it in a way that it is funny! Awesome!
His point about paying for things is really important!
Yes! Lanier has it right - Provenance of content - especially photos and audio/video recordings of real events, attested to on a blockchain (or functional equivalent) where anyone can see the original from which content is claimed to be derived AND verify that it matches a hash on the chain along with author info and their attestation of HOW the content was created.
The current idea of watermarking AI generated images is obviously doomed to fail from the starting gate, but if something is claimed to be a representation of a real event, we should be able to check that at least the author claims that and takes responsibility for the truth of that statement. And hold any fraudsters who make false claims accountable by penalty of law.
For really dangerous information (things authorities will want to keep hidden for example), authors should probably get digitally signed attestations by witnesses of the event represented in their content. Ideally in a way that witnesses can remain anonymous until they are called upon to verify their witnessing in a court.
A photo is never enough in a legal setting. Like all evidence it has to come with a chain of custody, i.e. we need to know how it came into existence and then into the possession of the court. That's already covered under current law.
the trouble with extreme provenance is that it is the antithesis of privacy. Do you want every single thing you ever did or said to be traceable back to you as part of a permanent record? Would you be willing to accept that anything you did online - anything - could come back to haunt you in 50 years time, in an unpredicted era of novel attitudes? It would be the end of personal freedom and of the 'right to be forgotten' - which is a little ironic, seeing as Jaron said he worked on GDPR.
@@RandomButBeautiful um... How do you get from "content creators record and certify their creations as done by them and attest to their veracity" to "every thing I ever do on the internets will be recorded"?
@@tomcraver9659 You are quoting me as quoting something that was never written. So long as you are fully at choice and do not lose your right to be forgotten, which is a real thing already encoded into the very law which Lanier has worked on..... do you understand?
Waltonchain has been working for several years on provenance in global product distribution. And then there are NFTs of course which can be used to create a unique signature for a digital file. Those things already exist. As to whether there is a way to track content and potentially block it from being used to train a model, that is only going to work if it is legally required for models to declare their sources. I don't see how any of them could be incentivized to take that burden on voluntarily.
@@RandomButBeautiful Sorry, I assumed you'd realize I was paraphrasing your comment, not quoting it, just as I paraphrased my own previous comment. You seem to be talking about a bunch of things that simply aren't related to my comment, which focused on solving the "fake content" issue, not issues of privacy or being forgotten or preventing the use of content to train models.
I absolutely adore this man.
Excellent - many thanks
Really liked the last question.
Criticism is the ultimate face of optimism. What a brilliant observation. So true and sadly actively avoided by most due to our modern day sensitivities. Would have liked him to delve more into the mathematical issue with open source creating hypernodes. Not sure i agree.
im frequently critical of many aspects of our society to the point that im seen as a "negative nancy" by others. hearing him equate criticism to the purist form of optimism made me learn something about myself. complacent people ("stop whining") are the real negative nancies, as they lack the imagination to see a better world than the one weve got.
Jaron's got the quirkiest mannerisms lol
Such a cool guy. We need more !
at time 17:23 providence is NOT the only way to detect fakes, also common sense, and face checking!
(which Jason seems to have little of) but in paintings and other things maybe the hardest and ultimate way to check
11:35 I agree with that actually, what the AI should be 'doing' IS the referencing. a) for attribution. b) to connect information.
Well said! ❤
the nicest gentleman in the cybernet
Brilliant interviewers!
Jaron as always on peak form. Love everything this guy says. But I really have to call out the presenters, they couldn't have been better at engaging with him!
I have *Great Respect* for Jaron Lanier ... As a pioneer of augmented and virtual reality filmmaking, Jaron was the OG.
This dude is quite insightful on a variety of topics
Jaron is a legit genius in this space. He's a VERY unheard voice that should be heard everywhere. Would love for him to get on Joe Rogan just to get some big big reach.
Jaron is a one of a kind brilliant human.
Its unfortunate to have a brilliant scientist being interviewed and only getting a small snippet of what he’s trying to say, the interviewers are veering back and forth between their pre-prepared questions without engaging the topics that he’s bringing to the table - would love to see the full raw unedited video.
2:06 What a great f****** question! Can’t wait to hear his answer! 💪✊🔥
great video!
He’s like a real life wizard …
Jaron is a hidden gem 💎
Huge fan got a occulas 2 because of him ...still learning!
4:55 in a perfect world. But I have a feeling these companies will STILL collect and exploit our data to the fullest even while we’re paying a subscription. Maximize profits.
I had dinner with Lanier back in 1986. He was the Godfather of VR. Crazy night.
Who would have thought you would hear Gang-slang such as 'O.G' on Bloomberg? 😂
When you consider the best thing about the domain of mindfulness and how we have so many people who try to make a difference for every single person in the world we still need a lot more than the people's who lose out of the technology and how it draws in a whole lot more revenue from the world but falls short of making the world better understanding the production center of economic development is important to consider