Post

Newsletter: There is no AI Revolution. Consumer adoption of generative AI outside of ChatGPT is barely 100 million people, every single company in generative AI is deeply unprofitable, Microsoft is pulling capex, and OpenAI spent $9bn in 2024 to lose $5bn. www.wheresyoured.at/wheres-the-m...

wheresyoured.atThere Is No AI RevolutionSoundtrack: Mack Glocky - Chasing Cars Last week, I spent a great deal of time and words framing the generative AI industry as a cynical con where OpenAI's Sam Altman and Anthropic's Dario Amodei hav...

1.5K reposts5.1K likes

Elderly here and started using Microsoft AI and like its conversational results but most have links to sites I would have gotten fr google. btw can't see past searches, can't cc results and have had wrong, bizarre answers . Not impressed

I have tried IA. It's rubbish. It doesn't get you what you want. You try again: the same output. You explain a little better: same output. You explain again, with more detail and clear demands about some items: same output. You explain step by step, you use 5000 words: same output. ALL of them.

I've learned to start search prompts with the word "fucking" to make AI results fucking go the hell away. All it's ever done is shown me lies so I'm in a state of permanent rejection of the entire technology now. It's worthless garbage, and will remain so for years to come.

Sorry to be lazy if this is easily available, but do you have something similar that focuses more on the false promise of the 'intelligence' side of it? Trying to de-influence a friend.

I'll check out the one linked article but I need more...my boy is a philosophy major living alone hanging out w Claude all day. Need a real intervention.

if there's "something" beyond being in the actual cult it's prob hubristic belief they can finagle irrational exuberance to accumulate fungible infra? they see nuts capital raises using the crypto/nft model, except w/ "ai" providing a fig leaf for it to be investment instead of gambling, & want in.

it would be interesting to see side-by-side analysis of this garbage with the timelines and financials of the og burn-it-til-you-earn-it VC-birthed golems of uber etc. - i suspect they'd look fake as fuck even in that context. probably somebody - probably you - has done this already & i missed it.

Is the money being burned on this only being spent on extra computer chips and electricity for the machines? Or are they also spending it on high wages for employees, various products, etc? Because it seems like these billions would otherwise be locked away in the dragon hoards of these corps to rot

This is all way above my head, but is this why OpenAI and company are looking to the federal govt for $$$? Like a VC with infinite pockets?

I wish MS would ditch it faster. I am so bored of trying to write things with Copilot going "Hey, you vant to bouy some toolips?" at me. Burst this annoying bubble already.

I think these people need to go away and think about things that humans are really bad at compared to machines, and make AI that does those things. Trying to make AI that communicates with other humans seems like the opposite of that.

It’s almost like firing all the people who would use GenAI so you can invest in building GenAI so you can sell it to all the people who would use GenAI was a—wait for it—not great idea.

I recently upgraded to a new phone. And just about the first thing I did was go through all the settings turning off, disabling or otherwise exorcising all the GenAI slop and bloat that came preloaded. I doubt I'm alone...

and it looks like he lied again, Grok is not 'open source' in the sense that you can download it to your computer like Deepseek or Llama, you can't, and there isn't a 'distilled' smaller version...not that anyone should want to use it anyway

Sorry, I meant if the AI bubble bursts because they fail to produce a product that creates a profit. Will they (like crypto and AI) just invent a new product to convince investors to pour money into for the next few years or will those companies fall apart and end this cycle?

Can't see why not. AI is just the latest string in the age old tradition to sell "the future" to many people with more money than smarts, in general terms. AI however is already leaving lots of collateral damage as is.

Ed, Is the play re AI that it can replace human labor and that they are just offering inexpensive entry-level fees to get people hooked and then will raise them once AI is seen as indispensable for saving labor costs, sort of a Jevons Paradox?

Won’t they just start focusing on selling AI to the military and oil companies? A vicious and profitable circle burn gas to power AI -> ai for military -> military maintains global hegemony over gas -> ai helps burn gas more efficiently -> profit and repeat

AI already took over. It's a new phase and it doesn't matter if the profits match the spread of it. I'm saying that because people have lost footing to the sheer popularity of AI

GEN AI ISNT ALL AI. GEN AI ISNT ALL AI!! Those thieving gypsies with their illiterate mechanical Turk are a side show! Look at Alphafold, look at a thousand verticals that will benefit from ASI. The revolution is happening, people are just distracted by the freakin' puppet show!!.

I wonder how many of those users are HS and uni students and their rates of use. There could be an alarming confluence of uneducated adoption, unaddressed COVID harms, & low buy-in/valuing of education that locks genAI into to social fabric via younger users. FBook, only worse.

That's my concern. I can't help but think the real reason for the LLM push is not to make a profit, but rather to create a generation of people wholly without critical thinking skills who will believe anything oligarchs tell them (indirectly through chatobots and social media).

Counterpoint: General Mills (GIS) says it is saving millions of dollars using AI to streamline its logistics process. I'm seeing that from other companies as well. I think AI is too diverse to make broad statements. Some things work; others don't.

Sometimes I wonder if the developers are blinded by the use of genAI for code generation, which is indeed very useful from my experience. There are also some promising enterprise applications. But that is probably nowhere near enough to justify the huge investments.

One thing I am not hopeful about is that if these people pull most of the plug on AI, they're just going to put money into somethings else that is also a net negative for society. They're not having any epiphanies about "making good products that are ethical and have practical purposes"

Interesting take that seems to only focus on the typical consumer product aspect. The unit economics do not make sense for the major players, but how long did it take the cloud providers to become profitable? If you don't see the value of GenAI and it's impact on the engineering community that's ...

something to look at. Simply put, we have unlocked something difficult to quantify, creativity in AI/ML related work. Software engineers aren't just following a product roadmap anymore. Creative freedom has been unlocked and people are finding it easier to create MVPs, and the GPU usage is declining

Much like the Blockchain early days, GPU costs were high but we're rapidly developing new models (SLM) and hardware to use less processing power and create less environmental impact. It's been a while since we've seen devs finding problems to solve, folks are enjoying learning, that's priceless.

You're practically speaking gibberish, and invoking blockchain to argue something is commercially viable is nonsense. The only people who made a business out of blockchain are crypto scammers.

GenAI as a standalone product is silly. Using GenAI to augment one's work to improve efficiency, quality, and scale certain functions... Now that's where companies improve their DORA metrics for faster time to market. Just I different take on things.

So? Corporate AI not making a profit doesn't mean the tech itself isn't revolutionary. Everyone who uses it just quickly starts running open source models. It's an industry that simply destroys the profits of every other industry, because it can do services nearly for free that used to be hard

A key part you seem to leave out when you cite the concerns of digital artists is that a great part of them center on the fact gen AI was trained on a lot of their works without permission. Also why its use is tied up in a growing amount of copyright cases, such as: www.wired.com/story/thomso...

wired.comThomson Reuters Wins First Major AI Copyright Case in the USThe Thomson Reuters decision has big implications for the battle between generative AI companies and rights holders.

Digital artists, copywriters, editors, junior programmers, search engines, StackOverflow... Soon SaaS, moviemakers, onlyfans - name the content and it can make it dirt cheap with tons of competition.

And you do realize AI video will be making blockbusters at 1/100000th the cost, right? Like, you can't seriously be seeing vids already fooling people's sense of reality for less than 10 bucks to render aren't going to massively disrupt the world?

The only time I’ve seen AI video content work is unsettling and creepy body horror videos and even then it required a lot of editing and work to produce ten second clips. AI video is a century away from being able to produce a Hallmark channel tier movie

"barely better than 4o", "imaginary product" - that's pretty rich. Is this what the AI skeptics believe, by never even actually using the tools? There's a very clear and very fast improvement with every model, far faster than any other industry, and escape velocity of "better than humans" nearing

Whether or not you believe AI will be adopted and demand will go up to make AI ventures profitable is not the point. The point is the tech itself is certainly racing forwards in capability, and will be capable of having its own impacts on the world regardless of whether the general public adopts it

At the very least, people should be treating these capabilities with respect and allocating appropriate risk to these improvements happening. (And as a senior programmer who has researched AI for years now, there really dont seem to be any scientific reasons it wont continue scaling)

In the past, plagiarism was relatively limited, so I'd usually make the student resubmit the work & move on. However, this is the first semester I've taught an online course, and AI use is rampant. I decided to actually follow university protocol & report them for academic misconduct (which I hate).

The "we are almost there" crowd, when talking about quantum computing absolutely never mention that of it happens, crypto as a whole goes down the toilet and loses their only "feature". Or if they choose to start using quantum computing for the chains, it becomes even less profitable. So... Stupid.

To be fair, 3D printing is absolutely still a Thing. It's just that, instead of marketing people rabidly trying to convince you that it can be used for everything ever, it's just being used for the stuff it's actually useful for. The same thing will probably happen with LLMs.

This also isn't like social media or online shopping with monopoly power where you get locked into big groups of followers and friends. There is no obvious monopoly power. Any new AI company can just pop in and undercut existing firms. There is no obvious way for AI companies to prevent that.

There must be a story behind why the UK is suddenly going full bore on AI. It’s as though the VCs are pressuring the government to ensure they get their $ back. I want to know if Starmer truly believes in the tech’s potential or is just bailing out the money men.

They are pot committed, and they have pot committed companies full of leadership who should not be in leadership positions. It's one big bullshit circlejerk. Talking smoothly with composure using jargon and buzzwords does not qualify you to lead. GET OUT Let someone who can do the job have it.

And that is exactly why there is such a push by some tech bros to use AI in our government. They’ll have lucrative government contracts to get a return on their investment. And, AI is only as non-biased as the people that code it, so we’ll be paying through the nose for a more unequal system.

The operating costs to run LLM’s never made sense for a general mass market consumer product. AI “agents” and chatbots deliver little tangible value. The real benefits are for data heavy scientific & industrial applications, but it’s hard to build a business worth $100’s of billions for that market

sure. Half of fortune 100s are using Blockchain at scale, just because you don’t know about it doesn’t mean it isn’t real this is how innovation works on the other side of the hype cycle. The same will happen with AI www.perplexity.ai/search/what-...

perplexity.aiWhat corporate 100s are using Blockchain at scaleMore than 50% of Fortune 100 companies are using blockchain at scale, with adoption spanning industries such as technology, finance, and retail. Key examples...

Should I dive in and read the AI drones telling you how mistaken you are by using tweets, press releases etc. from AI companies as proof of your poopyheaded wrongness. Or should I just remain ignorant in my Luddite hell.

The AI revolution is Elon stealing every last speck of govt data by querying AI models on his private servers to "look for" so-called wasteful spending within govt systems.

this captures perfectly how ive felt bout it the last couple years. the faith they have in gen ai has been confusing to the extreme and talking about it has been like screaming into the void.

I spent 30 years doing AI research, my specialty being natural language processing. I retired in 2016, thinking there could well be another AI Winter. I lived through the first one. To my surprise, along came generative AI, and the world seemingly went nuts.