Former Harvard CS Professor: "AI is improving exponentially and will replace most human programmers within 4-15 years:

Software creation will remain a rewarding and lucrative career. All that's changing is how we create the product. Don't pay too much attention to predictions from academics, kids. They have an almost perfect track record of being wrong. It's why they're professors.

The amount of years it will take AI to replace human grows exponentially. Last time i read something like this it was 1 year, before that it was 6 months :p

4-15 years LMAOOOOO, cast a wide net and you'll catch something

I predict that in 4-15 years it will be in another 4-15 years. Classic strategy.

Has this CS professor never heard of the curse of dimensionality or the brutal exponential nature of the scaled dot product? Fool, stick to teaching recursion and big O cause clearly those are the most complex topics his generation of coders went through

the ones who adapt will be the ones prompting circles around those still gatekeeping syntax knowledge

If you like this, you will enjoy our daily AI newsletter. Join thousands and learn about the latest breakthroughs in AI, plus step-by-step AI tutorials in just 5 minutes:

I thought CEO Anthropic said 6-12 month away, so who should we listen or trust? Hmm 😅 Me: Trust what I read from my code.

Maybe professors will be replaced since repetitive work is something that ai excels

In 4 year you will have developer on a completely other level. IF they use ai like a partner. Human are always needed maybe we should start coding guilds

lol cs professors have failed their students - they’ve been pumping out folks who never really understood what it took to build software and now they’re saying those students are cooked? 😅 Almost like the lack of real world experience has warped their world view.

3 months.

Yet we are sponsoring visa as fast as we can print , screw us citizens at employment opportunities

I haven't seen the full talk, but everytime somebody tries to compare AI to previous industrial revolutions and try to extrapolate what is going to happen from that I just cringe.

We're seeing the same view and prediction across the board, but believe equally as much that new jobs and roles will be produced because of the same tech advancements.

not having kids

Bro said 4-15 years

4 years for sure

It will not improve exponentially.d

Question is if it can replace the deployment stage, that deals with all the bugs that come when clients are using the software. Another question is if sufficient data will be available for training future LLMs as many programmers may take their hands off Open Source projects.

what kind of range is that? I can make predictions too! we will put based on mars within 5 to 100 years.

Dude works at Palantir, very credible source for those kinds of opinions 🙃

@grok who is this? And what are his incentives to say this?

AI demotivates and defunds the tourists, I’m ok with it

The technology they used to build these LLMs doesn't allow for this exponential jump, and money isn't enough anyways

@grok what is Matt Walsh involved in now that incentivizes this narrative?

Why the fk would you cut the video short for no reason?

6 months / 1 year later, someone will say same thing again! its a loop! or a ai-loop-hype..lol.

Ai is kind of coming to the same wall as quantum computing. The Ideas are space age but definitely not short term projection. Ai regurgitating a bunch of code it already knows based on probabilistic remixing is not a replacement for coders.

these Professor write books ..stop buying books , AI is enough to learn coding , ...Also i wish many students from India , Asia should stop spending 1000's of $ to attend US universities , AI is enough to learn .. overall stop spending money, AI will take care of everything 🙄

exponential for capability, slower for organizational adoption

How many programmers were there 3 years ago? I feel like I’m going crazy.