Is the AI bubble popping?

The story of 3 startups paints a portrait of an AI bubble that could be popping.

Subscribe to Synapse for deeply researched stories that are shaping the AI landscape.

CHAPTERS:
0:00 Intro
0:44 The Bubble
2:03 Stability AI
8:16 Stability: Key Points
9:41 Inflection AI
12:58 Inflection: Key Points
17:01 Cohere
18:37 The Big Question

#artificialintelligence #technology #startup #finance #business #story

emad mostaque,mustafa suleyman,cohere,inflection ai,stability ai,tech,ai,artificial technology,business,machine learning,bubble,funding,investment,invest,story,documentary,journalism,synapse,mannermind,entire story,ai news,artificial intelligence news,founder,startups,startup advice,open ai,openai,microsoft,aws,ai bubble,market crash,

[Shlooomth]

The real problem is that we’re still trying to do infinite exponential fucking capital growth instead of just actual science.

[nyx]

Investors are too eager to throw money at technology that they don't really understand. The training time, data collection, and energy requirements for these models is immense. Does it make business sense to spend $10 billion training and developing a model that will last for a few months before it becomes obsolete?

[ShazyShaze]

The marketing around AI is so frustrating to me. People want to focus on the exciting, sexy long-term fears of "rogue AI taking over the world, " rather than the boring legal and social implications of the technology. This often feels like a cynical marketing move to deflect attention away from the right-now problems of the technology: "Pay no attention to the serious issues with copyright infringement, harmful sexual content and social biases these AIs have. Ignore the severe limitations of the technology, that it's not actually intelligence, just a hyper-advanced pattern duplicator that we don't even fully understand yet. This tech is actually infinitely powerful, and that's super scary because it might SkyNet us, and you need to give me and my company a ton of money so we can save you from the Terminator future."
The tech is extremely impressive, don't get me wrong. I just think it's exciting in a much less sexy way. It has limited use in some bulk work pattern recognition tasks, and it's fun to generate silly images and such. Worse, it has issues that need to be addressed today that seem to be disregarded by the folks at the helm of this stuff.

[annaig]

About a year ago I got rejected from a program at my school (comparable to a short-term internship being held in New York) because during my interview a question surrounding AI was brought up and I discussed how it was similar to the Dot Com crisis. I went on to support the idea of an AI Bubble, but my interviewer called me "too bearish towards Artificial Intelligence". I know he probably doesn't remember me, but watching things only escalate since then makes me wonder how many others are still blindly following these startups..

[wdmeister]

It's like with crypto. New "AI" firms are popping out left and right but they don't do anything more than what you can do in ChatGPT, Claude, or LLama but get crazy valuations. 99% of "AI" companies are like "let's use OpenAI API create a few simple agents and pretend that we've made something unique".

[NonCompete]

I hate when people use average wealth growth to talk about the economy. The truth is that wealth inequality is so off the scale that averaging out wealth doesn't make any sense. More than half of Americans can't afford rent right now and quality of life is very clearly deteriorating for working Americans as inflation soars. Americans aren't getting richer in real terms and it's obvious to the hundreds of millions of suffering workers getting gaslit by this kind of propaganda, being told their lives are improving as they're actively falling apart.

[nilaier]

To be honest, people still use Stability's models due to their open-source, uncensored and customisable nature. So, even if Stability wasn't able to make their product profitable, they definitely made a big impact on AI landscape.

[fishraposo]

As someone whose job title is literally "Ai Specialist", I'll say yes. People are blowing it WAY out of proportion and it's already backfiring.

[Psychojogia]

I was wondering how one can pour tens of millions of dollars into a free-of-charge app... now I know: one cannot.

[szebike]

The main problem for any current AI system is realiability. They can do some fun hobby+ level assistant things but when it comes to important, scientific stuff where accuracy is needed its too unpredictable und unreliable. You can't monetize on hobbyists or low level ad makers. To make them scientific reliable you basically need AGI level AI which is decades off I think.

[rakeshbarman]

Once Isaac Newton said ' could calculate the motions of the heavenly bodies, but not the madness of the people'. Once again same is happening in AI SECTOR'

[danheidel]

Back in 1998, I took my last college undergrad course and it was a lecture discussion about the history of technology. It was a really fun discussion group. The take home from looking at informational tech advances since the Gutenberg printing press was that the short term impact is always massively overstated and that the long term impact is always understated and vastly different from what the initial hype promises.

[usernamehandle]

That "typical American is 100x richer than 1870" quote is so incredibly skewed and inaccurate, I am struggling to take the rest of this video seriously.

[miss.dudeman]

I feel like the problem with the ai bubble is, nobody really wants this boom in products and promises of what ai can do. It has more so sounds like a detriment to humanity than a net benefit. It’s like their creating products in search of a problem, and causes more problems than they solve. Before when companies used ai, it was quite and just another part of the process. The ai product that keep on being pushed out by every single company, just to make their shareholders happy, are just loud and anoying more than anything.

[lasura]

Generative AI is NFTs all over again. When a tech business tries to push into art, you know they have no real use case.

[a_ghoul]

I disagree with revenue growth making us all richer. It will make the rich richer like has always happened especially in big tech

[ShumaBot]

Hopefully this signals the collapse of this entire industry of thieves and conmen.

[LokiScarletWasHere]

One problem that's making the AI market expensive is nobody's *coding* AI anymore. Nobody's *making* AI anymore. Everyone's just basing everything on existing models, nobody's trying new methods, the difference between each one is mostly just the dataset and the epochs of training it's been through.

[cencent]

The average amarrican would not be 1000 times richer. Not everyone will benefit from that growth

[Badwolf]

Tbh i watched a video not long ago about some artist losing his job to AI 6 days ago as of writing this. I get so many people saying that AI will be used along side people, or it will take years for AI to fully develop. And I realised all these people are saying different opinions because in reality, nobody knows. But knowing that some people have already lost their jobs to AI worries me. Because above all, it's happening now.