The AI crash is impossible - Change my View

[bazera]

3 hours ago, Natasha Tori Maru said:

I don't think AI is going to take anyone's job outright - but I think the person who knows how to use AI will.

For sure.

That forces us to learn to use AI tools as employees, especially in programming. And that can be demotivating sometimes because you were good at your job for 10 years and now somebody with 3 years of experience can do more with AI than you can do manually, and you are forced to jump into AI bandwagon. Forced to join AI rat race.

Sometimes I wish we could go back. 

But maybe people also thought like that during transition from horses to cars, who knows.

Only time will tell if the effects are positive or negative.

[bazera]

Lol

[LordFall]

 

[NewKidOnTheBlock]

It's kind of amusing how Leo is able to entertain and just internalize any crazy possibility of what might happen, or what might already be the case, out there; but A.I. replacing us all pretty soon? NEVER LoL such an unthinkable prospect. Because it's not like there's a global race for AGI and there are massive incentives to just keep going and accelerating the research and infrastructure building, and it's not like the demand for A.I. is essentially limitless right?

[Lyubov]

22 hours ago, NewKidOnTheBlock said:

It's kind of amusing how Leo is able to entertain and just internalize any crazy possibility of what might happen, or what might already be the case, out there; but A.I. replacing us all pretty soon? NEVER LoL such an unthinkable prospect. Because it's not like there's a global race for AGI and there are massive incentives to just keep going and accelerating the research and infrastructure building, and it's not like the demand for A.I. is essentially limitless right?

Dude the moment you are actually  employed and you integrate it into your workflow you will see how much of this shit is hype and speculation. It is not going to substantially replace everyone. It’s modestly replacing some juniors and making some people more efficient. The layoffs aren’t because of AI. They’re genuine economic downturn and the erosion of the middle class with several problems arriving at once. Silicon Valley is just riding off that and reframing everything as “it’s AI” so people dump their life savings into investing in it. It’s actually very convenient for them from a business/ hype perspective. I’ve noticed a large uptick in quality YouTube content on people starting to share their doubts and issues with AI, obviously it’s more nuanced than forum posts here. 

Edited yesterday at 09:37 AM by Lyubov

[NewKidOnTheBlock]

23 minutes ago, Lyubov said:

Dude the moment you are actually  employed and you integrate it into your workflow you will see how much of this shit is hype and speculation. It is not going to substantially replace everyone. It’s modestly replacing some juniors and making some people more efficient. The layoffs aren’t because of AI. They’re genuine economic downturn and the erosion of the middle class with several problems arriving at once. Silicon Valley is just riding off that and reframing everything as “it’s AI” so people dump their life savings into investing in it. It’s actually very convenient for them from a business/ hype perspective. I’ve noticed a large uptick in quality YouTube content on people starting to share their doubts and issues with AI, obviously it’s more nuanced than forum posts here. 

No, I think the situation is actually more nuanced than YOU and many of these YouTube commentators talking about bubbles (most of whom know nothing about AI or programming, nor do they have much analytical and problem solving skills) portray it. AI ITSELF is NOT actually a bubble, it’s just that some companies, mostly OpenAI, are having some existential troubles due to bad strategic decisions and the way they communicate with the public. The economic downturn, bubble, and scam allegations are just a lazy cop-out; a form of cope for luddites and people who fear change. Worker replacement is also more nuanced than how you described it, and the main thing blocking it is - and always will be - companies' and people's unwillingness to adopt AI. It’s not that the AI itself is incapable of replacing humans or that it’s less effective.

We are currently in a transitional, pre-adoption period from what I can tell. Research is moving much faster than implementation and the people's willingness to implement it can keep up with. But it's going to happen eventually.

I personally knew programmers who are already using AI for everything, and they’ve been telling me it’s just a matter of time. I think I'll take their word for it LOL AI is not the 17th century tulip bubble, it’s not really even the internet boom/bubble. AI is legit Industrial Revolution 2.0. I think people are forgetting that we’re dealing with a completely different beast here. But we'll see who's right and wrong. Or maybe we won’t.

[AerisVahnEphelia]

it's going to correct, but not crash.

but maybe it won't.

I keep seeing improvement, it doesn't plateau, but they are slower, and the more it improve, the more smart you need to be to see the upside, they are not easily accessible anymore.

[NewKidOnTheBlock]

4 hours ago, AerisVahnEphelia said:

I keep seeing improvement, it doesn't plateau, but they are slower

Not according to these graphs. In fact, the progress seems to be accelarating as per expectations

https://epoch.ai/data-insights/ai-capabilities-progress-has-sped-up

https://epoch.ai/benchmarks/eci

https://epoch.ai/benchmarks/metr-time-horizons

https://epoch.ai/trends

https://epochai.substack.com/p/the-epoch-brief-february-2026

 

[AerisVahnEphelia]

1 hour ago, NewKidOnTheBlock said:

Not according to these graphs. In fact, the progress seems to be accelarating as per expectations

https://epoch.ai/data-insights/ai-capabilities-progress-has-sped-up

https://epoch.ai/benchmarks/eci

https://epoch.ai/benchmarks/metr-time-horizons

https://epoch.ai/trends

https://epochai.substack.com/p/the-epoch-brief-february-2026

 

Measure is relative, but when I say slower . . 

[LordFall]

People have different views on the topics but I would seriously break down what they're saying. The experts interviewd in this piece for example seem to not really have thought out what they're saying which is wild because Arvind Narayanan should have the background to fact-check himself. 

He talks about reliability vs capability of models but models can be set on a setting to lower their variance and just put out consistent information instead of the greater variability that commercial models like ChatGPT and Gemini thrive on.

He talks about predicting that unemployment numbers will remain unaffected by AI in the coming years when it's already increasing more than planned. 

https://www.cnn.com/2026/03/06/economy/us-jobs-report-february

There is a girl working at OpenEvidence a new healthcare AI startup saying that AI should be held responsible for false diagnosis. It absolutely should and so should the physicians working today. Anyone that's had experience getting sick or having a sick family member knows all too well that you go to 5 different doctors and often you'll get 5 different diagnosis or treatments suggestions! 

Arvind Narayanan closes out the piece by saying that AI will not be able to predict the future as it's too complicated, no shit cowboy, hasn't stopped humans from trying it from the dawn of time and a lot of them are correct. And AI is built to process data and make suggestions.

A lot of these discussions are getting me to see the limits of human intelligence and even the top experts in their field use thought processes that make no sense and are provably false. I'm very excited to have AI intelligence take a hold of our society and raise the bar higher than the educated apes we are today. 

[NewKidOnTheBlock]

Agreed. Human emotions and biases get it the way of accomplishing things way too much, in fact all the time. I've started noticing and realizing recently just how many human dialog/conversation trees effectively serve no function whatsoever. This should have been obvious to me a long time ago but I wasn't attuned to it because I'm also just a human with my own inneficiencies.  A.I. might sometimes or even frequently halucinate/make stuff up and it's limited by it's lack of agency and the lack of ability to genuinely experience, but the inteligence differences are there already. It's only gonna get more and more evident as the time passes