Post

Dear Ed, I vehemently share your position on 'AI' and your frustration, too. Keep speaking the truth. You're not talking to an empty room, as far as I'm concerned.

when the bubble bursts… bsky.app/profile/cate...

I've got it...genAI and AGI is like cold fusion...always a decade off...

Popularity of "AI GPUs" certainly affects gaming gpu pricing, particularly the high vram 90 tier

This is the biggest bubble since blockchain. And the same people are trying to make that a thing, despite the fact that no legal use case has been found.

Whether OpenAI will be profitable is up for debate. Whether LLMs are very powerful and useful technology is not

can't say he didn't warn us 😂

Thank you for writing this. You are not alone. GenAI is a bubble, capitalism's last hope in these dire times. Maybe its the last bubble ever, who knows 🤞

Ok

I think a lot of the SV companies don't acknowledge that a lot of what allowed the hypergrowth markets were the conditions of the post-2008 crash. Interest was low, and [Democratic] regulators did not get in the way of all the issues because they offered growth and GDP increases in a rough time

It’s not there yet, and it may never be there.

so I just heard ChatGPT is trying to pivot toward acting as a shopping platform and that seems like the beginning of the end tbh

Every earnings call for every business seems to feature some attempt to inflate valuation using 'AI solutions' even when it's just run of the mill automation. I'm not sure how bad a bubble burst might be. I was a kid when dotcom hit so I don't have a real reference.

I read that in your voice 😭

youtu.be/UoquS5uXOKs?...

This is spectacular

Well maybe they wouldn’t be if “generative AI” companies actually paid them to use their work. But they don’t.

I just want to say thank you for writing such a clear and concise, comprehensive explanation of all the things I've been so angry about ever since this grift began. It was refreshing as hell to see facts and righteous anger combined into an absolutely damning expose of it all.

“There are no other hypergrowth markets left in tech.” the VC shops (that actually had good returns) closed up just before cloud craze as the writing was on the wall already :^)

He is also wearing rainbow suspenders.

You may be the greatest writer of our generation, man.

Valley - maybe the whole tech industry - into a Black Mirror version of The Emperor's New Clothes. Now, a lot of influential people desperately *need* it to be true. A whole lot of manifesting going on. As it has done before, the industry built on hope. At some point, awash in $, hope became a lie.

Their benchmarks are also testing their own products which are being developed with minimal, lip service transparency. Strong "trust me bro" situation.

I'm tired of fielding the same half-assed critiques of my work, so I've condensed my answers into one helpful section - no, the cost of inference isn't going down, no, OpenAI isn't growth stage, no, it isn't like the early days of the internet. Come at me. www.wheresyoured.at/reality-check/

On the cost of inference going down: LLMs seem to have hit a wall in terms of what they can learn (they are already trained on the whole internet). If we move to models which are capable of learning as they go, then inference and learning become entwined and the cost will sky-rocket.

Welcome to the economy of companies that can't turn a profit but are valued at astronomical valuations. We need more leaders/managers who can add value to customers while making a profit for their organization. If this disjointed economy continues, we are in for a rude awakening, circa 2008.

Yes, and no! I was on the OpenAI private beta, prototyping RAG, tuned to their ethical standards, and OK'd for launch. Then they launched. Out went the ethics at the first whiff of cash, and in came a second AI winter of distrust for independents like myself, for whom now spare a thought!

So you get headlines like "RFK says measels vaccine is ineffective", and then the article is just stenography of comments he made. MAYBE if you're lucky, an expert in the 2nd last paragraph says it's wrong, but even then the article won't take a side. Just platforming bullshit as if its news.

I reckon agents could be a huge revenue source. If users are willing to install a closed source application that they give full permission to access all their data and bank accounts I can think of a way that could make a lot of money very quickly, no LLM required, except as window dressing.

For comparison, Apple's revenue bumped from 20b in 2006 to 26b in 2007 then 34b in 2008. Almost doubling in two years on the back of releasing the FUCKING IPHONE in 2007. This was a meteoric rise the likes of which many companies could only dream.

That’s why I decided yesterday not to buy a new MacBook after all. The new ones are full of AI. Decided to keep my old one.

That chart is bonkers. Their revenue is going to triple this year (and then double the next year?) I accept, reluctantly, that there ARE people out there who actually want to use this stuff. But aren't most of them already using it?

Wiki article for more context en.wikipedia.org/wiki/Gell-Ma...

Been laughing like an idiot at "Big winky on the skillet bowl" for a couple minutes

💯

incredible writing. you've tried to be the reasonable skeptic for long enough that the whip brandishing is def in order

u absolute fools, this was a trap. a trick. a honeypot of sorts and u people in the replies fell for it.

I feel like there's still a distinction there. Sensors, object recognition and predictive path finding *could* solve the self-driving problem. They're just stupidly optimistic with when. "Stringing words together based on likely occurrences" only solves "create plausible but often wrong text"

We love you Ed

📌

I sometimes say "consumer software industry," but nobody knows what I mean...

The media thinks the tech beat is like entertainment and sports, which are basically free publicity for entertainment studios and sports leagues. Until recently it was, but not anymore.

"The ER didn't get me in to see the right specialist, now I'm waiting forever" Man, that is not what an ER does. And have you been to a hospital or seen the state of healthcare in the US? AI ain't fixing all that.

Why waste money on a sever farm when OpenAi is already on the way to bankruptcy on it's own?

One time someone accidentally gave them $1.00 in sticky pennies

Adding, as a resident of Utah and one who has morbid interest in MLMs and other, uh, "interesting" businesses I have to at least kind of admire how much OpenAI is able to raise money off of hot air.

What, putting air on top of uncertainty and wrapping it in a crap sandwich doesn't make money??

We just need AGI and ASI bro!

It won’t go away forever. It’s been here for decades already.

Is there some sort of hidden play where openAI takes a bunch of ad money away from Google and Meta? Otherwise, I don’t see how they hit these revenue wishcasts

thank you Ed Zitron...I was believing it all until I started reading what you say..........

📌

The year is 1994, replace every "AI" occurence with "Internet".

I thought the grift after NFTs couldn't be worst/more environmentally destructive + pointless, but I was wrong. But hey, maybe the NEXT pump-n-dump will be a little more ethical 🤞

I have only just started reading this and I'm already getting Enron vibes. Looking forward to reading the rest. Thank you for breaking things down like this, I don't know enough about tech or finances myself to articulate why it seems like such a scam.

It will be like Internet sites in the eraly 2000s and virtual worlds, it's going to go through a heavy winter and down cycle. Awful lot of money burned on nonsense.

at hank green.

get 'em ed

Generative AI will suffer the same fate as cryptocurrencies and NFTs.

god she was so iconic

Her one mistake was picking a field that had actual standards and stakes. If she'd promised, say, car diagnostics based on an oil sample, she'd still be grifting like a motherfucker.

i know, none of them are smart, it is just a grand lottery, they just play often and have means to do so, all they need is just one big win out of 20 failures

That’s how I felt today when wired posted an article about how an “acclaimed” book no one’s heard of was written by ai, and then in the article the author of the book (an Italian philosophy teacher pretending to be Chinese) says ai didn’t write it. And then the article just continues on as if it did

I don’t. Maybe that’s something

Yeah, at the time I could train a simplified FSD model that could handle limited environments on pretty basic hardware and run models on an RC car. The main hurdle appeared to be edge cases and liability questions. AGI doesn't have any demo and can't run models on desktop hardware realistically.

(bitterly) laughed my way through this one, extra spicy and chock full of great grabs like these thx again for your hard work, covering this shit must be a huge pain in the ass

Much like the dotcom bubble in that the technology is very real but the hype has been monetized into wildly unrealistic promises. Hopefully, this buries the SV VC cult, and the people actually building something can get back to doing the work.

Thing is the underlying tech is a decent advancement in the ability to process data. If the industry had just been honest about it, everyone involved would get a pat on the back and we could have enjoyed some modest performance improvements and new features. But they had to say they invented god.

Again, is Deepseek the final goal of AI? Why would you copy something what is not doing the thing, what AI claims to do at the end.

But Microsoft has pulled back on the compute a lot?

The fact that Sam Altman uses vocal fry like Elizabeth Holmes tells you all you need to know about Sam Altman and any company he is attached to. I mean, if the guy lies about his own voice then you really don't want to buy anything he sells because its all woo-woo snake oil bollocks.

A kitten dies everytime someone posts AI slop.

I have felt like I’m going mad while everyone points gleefully to their dumb action figure selves and crows about gen AI - thank you for writing out what seems so obvious but apparently is not

once again ed dares speak truth to power or at least tech marketing pipe dreams

📌

📌

Nonsense doomerism im afraid

how would they do that?