Microsoft AI CEO forecasts human-level AI in 18 months | Hacker News Reader

Microsoft AI CEO forecasts human-level AI in 18 months

twitter.com

13 pointsmomentmaker20 comments

    > However, as AI experts hypothesize about when, and if, AI will disrupt white-collar work, the technology thus far has made only a small splash in professional services. [...] Any returns the economy is seeing are largely confined to the tech industry, suggesting that AI disruption has been limited in the real economy.

There's one quote from the Dan Olson's "Line Goes Up" video essay [0] that's been rattling around in my head for a while now. (Predictably from 2022, if that was not obvious)

    > [...] it outlines just how disconnected from reality the people actually building cryptocurrencies really are. They don’t understand anything about the ecosystems they’re trying to disrupt, [...] and assume that because they understand one very complicated thing: programming with cryptography; that all other complicated things must be lesser in complexity and naturally lower in the hierarchy of reality. Nails easily driven by the hammer that they have created.

Specifically him calling the attitude the "technofetishistic egotism of assuming that programmers are uniquely suited to solve society’s problems" has stuck with me. I don't think AI developers are good judges of what a lawyer does in this case, so I'd rather be skeptical. I'm not saying that LLM tools AREN'T useful to lawyers or are capable of doing lawyer-like things... I'm just willing to humbly assume that I do not know what makes a good lawyer.

[0] https://www.youtube.com/watch?v=YQ_xWvX1n9g&t=1405s (timestamped to quote)

Why link to twitter post about a Fortune article? Just link to the article.

https://fortune.com/article/why-microsoft-ai-chief-mustafa-s...

“We know how to build AGI by 2025” - Sam Altman

Its all BS unless you have a demo to go with your remarks.

I feel like ChatGPT is general and intelligent. No?

It's not intelligent.

I could theoretically do the same math on a piece of paper to generate a list of tokens, then decode all the tokens into a sentence and read the sentence I generated for the first time at the very end of the process.

In this process, where was the intelligence? It didn't come from me. I didn't know what sentence I was generating until I was able to read it at the end. Was it in the pencil? The paper?

Intelligence is doing the right thing at the right time. Its an ordinary process that takes space, time and energy.

Demos are also often misleading and cherry-picked. Using AI to do one cool demo that breaks down 99% of the time when circumstances slightly change has played an outsized part in most of the AI insanity we are living with.

Didn’t they tell us this was going to happen several years ago? These predictions are embarrassing at this point. This is also a terrible way to garner public support and does nothing but strengthen opposition to these AI tools.

Keep promising until it eventually happens! Infinite money glitch.

People in tech think everything is changing, not realising that it is mostly the tech industry itself that is getting impacted while every other industry remains relatively the same as it was.

The sun rises in the east, so we turn in that direction. Bellwether track record for tech industry is formidable.

The "version of this story" from February 13th, 2026 as noted in the article was originally posted here:

https://news.ycombinator.com/item?id=47006000

Cursory look on wayback and it looks pretty similar to me.

Maybe easier to just republish the article every few months vs making a new claim and doing a whole new story?

so he will be out of a job in 18 months?

The way the headline reads the CEO is AI and it forecasts being at human level in a year and a half.

I didn't pick that up until you mentioned it. Lol. Nice one.

Reminder that Google, who literally invented the transformer, say that's never happening:

https://xcancel.com/Hesamation/status/2045181640297578605?s=...

I'm more inclined to believe that over someone who failed to understand that Windows Notepad didn't need AI.

I am inclined to believe we will someday achieve impressive AI, but I am starting to believe we will need more software engineers to “break bread” and work with Data Scientists, who are insanely inefficient and it shows, repeatedly. All their AI breakthroughs are always old computer engineering lessons.

There’s something deeper you are not pointing out.

Where are the Einstein et al equivalents of the modern age? I highly doubt they’ll ever come about.

We will continue to use old ideas as leverage but I would be surprised if novel stuff comes from the current generation. The oldies were built different.

I wouldn't be surprised if they exist but they get scoffed at, mocked, and completely ignored due to their ideas not being peer reviewed and outside of the known norms. How is someone with a revolutionary idea supposed to bust through all the noise and be taken seriously?

Google is not a person and this is not of one the authors of the transformers paper. Many are also already aware new ideas are needed beyond transformers.

Reply on news.ycombinator.com