Two Computer Scientists Debunk A.I. Hype with Arvind Narayanan and Sayash Kapoor - 281

@TheAdamConover I would love it if your videos here had some sort of link to a works cited page. You're a great initial source of information, but when I'm formulating a written argument for one of my classes, I'd like to reference and trace back the sources and data myself as a matter of due diligence.

Great video, thanks!

The golden age of software developers is coming to an end 😢

This coming from the same guy who got "ruined" by Joe Rogan 😂

Scapegoat

The best thing about AI is that it doesn't exist. The approach is completely wrong. If you want true AI, you need to start with natural intelligence and enhance it.

Ads... ads... ads....god no.

Fools. AI is going to be the end humanity & the animal kingdom much sooner than many think.

Self-driving cars being years away is a horribly ignorant & irrelevant analogy. Waymo is ALL over LA NOW!

AI will be the excuse for what happens, child molesters run the world and when you censor speech you harm a child if the if there was no censoring of speech we would know the truth, thank you Google

Investment-Baiting is a term I've started to use to describe this phenomenon. When companies seek to illegitimately inflate valuation by using poorly understood principles or technology in order to gain capital investment.

Jeezus, I feel like I'm at a monday morning standup with these two

nope

Very clear-minded guests. One bone to pick …. “You don’t have to understand the math in order to develop an informed opinion ….” True-ish, but … you DO have to understand that it IS driven by math (not magic) and that in fact it is driven by a branch of math dealing in PROBABILITY. It’s misinformative to suggest NO inkling (aka education, research, reading) is required. Not having any inkling into how the sausage is made just leads to fruitless and often counterproductive speculation. Everybody read something like “Why Machines Learn” at LEAST.

Dude chat GPT continually makes mistakes identifying animal species. But if you don't double check. You're being scammed.

Why don't you be honest about the inefficiency of deleteme?

I'm tired of seeing these apocalyptic videos of people saying how AI is going to take over the world and throw us into technofeudalistic society.... Like we're already in one. There's no reason to throw more fear into people's hearts making them feel as if they can't do anything about it.

Agreed with almost all of the video, but I think it is a bit disingenous to think that capitalism will make the situation better because companies have to hire outside their original hiring pool when they get big, or because he had a refreshing conversation with a hedge fund guy. These generative AI models are consuming vasts amounts of energy and CEOs will absolutely cause damage to the environment if it will possibly make profit to shareholders.

I love these videos, but you referencing Foundation...my guy...I have a ton more respect for you. Those are just such good books, and if anyone on this feed hasn't read then I suggest it. It is dense but so worth it. There are 3 books I believe.

Your appeal to the audience for the company, "Deleteme" was a very good advertisement. I will be checking them out.

I wish there was more effort to describe what a company actually produces and its value to me as the consumer than is normally the case in advertising. You did that.

i tried to find more about Arvind Narayanan, but after this video and others critical of AI he just disappeared into thin air. No blogposts, no podcasts, no socials. Very very strange

Most of these conclusion common sense should lead people to if they take the time to think. But I will say it funny how this ends with go use the technology from two people in industry. Lastly, Daniel O’Connor has some good thought on how this could be used from a perspective of a Catholic who actually believes prophecy anyway (CCC 675 +- but if you use the AI to avoid reading things yourself likely won’t mean anything).

I love how he found minority opinion computer scientists to prove his unscientific point.

Hey. ChatGPT, will the Magic 8-Ball be successful in the marketplace? ... "Definately, Yes!"

Not to the mathematical logicians and mathematicians. Maths people know the theory of automata and foundation of maths. Engineers however do not understand the subject so he make stupid claims.

These were incredible guests. I learned so much, and I really appreciated their message about trusting our own critical thinking skills rather than just deferring to “AI experts” as we think about the effects of this new technology

@TheAdamConover i dont know if this will work, im a little concerned that despite being om our side amd all, i find the fact that i cant get ahold of you to share something that is absolutely world changing, kjnd of dissapointing. Because well, it just feels wrong. Are you Us? Or Them. I can talk to Us whenever i like, we aremt too important or busy. But them? Nope we arent allowed to bother them. Butbright nowni feel like that. And indo have something ridiculously important to tell the world, amd ive decided i need to tell you first, even thouhh you wont believe me. Cmon be a part of n us, and hear me out

So we've actually invented an answering machine 😅 i missed those

Also, some of these things are unneeded solutions. For example on the self-driving cars. Look at thr Netherlands, Norway, Sweden, Denmark, Paris in France, and Tokyo in Japan. They have very low or even nonexistent traffic fatalities by prioritizing transit, walkability, and cycleability by filtering cars onto only certain through streets and only building freeways outside of the urban core while slowing down and preventing through running any auto traffic that is not on a major throughway while also making major throughways paralleled by safe and quiet walking and cycle paths, and cross able by the same and also paralleled and easily crossed by transit. Sometimes, the problem is systemic and doesn't need a new computer technology or a technical solution that is still in development it needs systems and infrastructure engineering, design, and minor social engineering.

Finally some real people who know how to tell people the truth and not marketing hype. AI is dumb as a chipmunk. All products today are crap but monry has to move so nothing but crap projects. Even gpt is crap with looping answers and crap code. It will take so long to make AI actually a replacement to experts and people. Simple tasks and repetitive tasks yes which are predictable but the rest no.

Like, do you live in the US? Its big. Many cities. Many layers of government. Much commerce through trucking.

We as a Global Civilization are well past the benefit of further software upgrades. We have gotten to the point of diminishing returns for quite a while now.

Our focus needs to be on physical technology - advancements in Architecture & Engineering that works in harmony with nature - and how we live to promote the best human experience.

AI isn’t it.

Addition of AI features may encourage churn between products - it's the new "hotness" - but ultimately people will figure out the results that matter to their bottom line are no better, if not worse, than human driven results. The things that AI might excel at are mostly low quality low margin content.

IMO what is really going on here is tech companies are using "AI" as the perfect cover to slash their work forces and boost profits temporarily. This is great cover for yet another lackluster growth year. Yay, higher profits! But wait, so many middle income people are out of work now no one can afford our overpriced products. Meanwhile companies they are locking themselves into being dependent on AI providers who will then consolidate and gouge their customers for those AI services. Unless something radically better comes long in the next year or so we'll have a bunch of shitty AI based products all competing with low margins and ever decreasing product quality.

Nah deepseek AI has been amazing for me lately whenever i have botanical or mycological inquiries that are really specific. And with culinary questions or stuff pertaining to fermentation. I would venture so far as to say that it "actually works"

It's getting old to see clueless execs continually work themselves into a tizzy over each AI buzzword that emerges. They use their scant AI knowledge as a dick measuring contest with their other execs. "Ooh, look at me, I'm a big suit wearing boy who just saw the term 'Agentic AI' in a Forbes article I didn't finish and now I'm going to make the peons look into it." "Have you heard about Notebook.lm? Aren't I so great for knowing about it?" "I'm going to sign up for a predatory meeting summary product like Read or Otter or even both and have an uninvited AI bot join every meeting I'm invited to and then blast all participants with an automated shitty summary email and then brag about how great my shitty predatory bot is as if everyone doesn't already know about it."

"Predicting whether a person is about to commit a crime" is not thought police, it is a normal part of police work - prevention cases, as opposed to investigation cases. A glorified excel sheet is exactly the right tool for assisting with that part of caseload: these cases are research-heavy and are typically more time-consuming and exhausting than investigations, precisely because there is too much data to go through for a human. "AI" that is shouldering work better suited for a machine than a human is very, very helpful here. It being a dumb parrot at its core is a feature not a bug, because in that space nobody expects a robot to actually assess the results in a comprehensive manner that requires human experience and intuition on top of data. But grunt work that's perfectly within its limits? Oh yes please, dumb baby skynet may even get a donut for its efforts.

Ai is now mlm. Even uf you see ai news channels just by their faces you can see they are nothing but scammers creating hyping even if just a normal ai update releases and other' losers watch them

🤣🤣🤣Great conversation! Any C++programmer can write an integrated algorithmic sequence and call it AI and the public would never know. Look there are some fascinating chips being made, so who knows what the future holds.

I find ot funny that literally the whole country DOES NOT WANT AI becouse of the danger, yet all these money bags want it. This has convinced everyone that these idiots really should get power drill brain surgery

is just initial stage. in a few years just face ITS NOT HYPE.

So far all I have discovered is that AI is just a sofisticated Google search lol

I just love your show, i saw it only now , out of date, and your are yet all correct , thanks and good job, congrats

This is a real video right?

These guys dont stand in front of rows and rows of books to boost their authority....

If only the world was run by sane techniocrtats as Adam and his chums seem to think.
Alas, whilst the engineers know the limits of "AI", almost everyone in Business, Investment and Politics doesn't give a stuff about this - For them, "AI" is a magic word that - for the Businessmen means 'Layoffs so more profits' , for the Investors, it's 'Magic Money' and for the Poiticians (the ones that AREN'T in it for the idealism - so, that['s quite a lot of them) it's "The thing my advisors say we must 'Embrace' ".
UNLIKE "AI" there's no actual uncertainty about the catastrophic damage being caused by Climate Change and the role of Oil in Party Politics and our Media....
And yet the most clearly insane laws and policies have been carried out for the last 40 years which not only doom but (more relevantly to the market) impoverish us all (even the 0.01%) ... and yet that's not even mentioned as an issue outside of some rarefied "academic discussion".

We just got Greed and Hatrted made into the ruling party of the USA ...


No doubt the Engineers will point out to us that this isn't possible because it's plain stupid and suicidal.
Remember when you got bullied at school by stupid people?
How did "This is illogical" work out then?

Yes, the problem is not the machines, it's the people who own them - and us.
And you KNOW they'd rather make Game Shows than Drama.

One of my biggest gripes with the entire AGI thing is they don't think AGI is AGI they think it's just AI but more