2026-01-14 07:31:32

Regarding the timeline for AGI, industry experts generally expect a breakthrough around 2026. A more aggressive prediction is that once artificial general intelligence is achieved, subsequent performance growth through algorithm iteration and self-optimization could reach a thousandfold or even ten-thousandfold increase—this logical deduction is actually quite conservative.

Interestingly, the idea that AI's comprehensive capabilities will surpass the total of all humans before 2030 was considered science fiction a few years ago. Now, this forecast window seems to be getting closer. Want to keep up with this wave of technological advancement? Many leading tech companies have launched free basic AI courses covering the entire chain from principles to applications, which is a great way to get started and deepen your understanding.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

15 Likes

Reward
15
3
Repost
Share

Comment

0/400

GasFeeCrier

· 01-14 08:00

2026? Honestly, I have some doubts, but what if it actually happens... A thousand times, ten thousand times growth, I didn't expect that, isn't this playing with fire? Surpassing all of humanity by 2030? A few years ago, saying this would have gotten me roasted, but now it’s kind of exciting. But those free courses all seem like schemes to harvest the leeks, right?

View OriginalReply0

DefiPlaybook

· 01-14 07:44

According to data, the consensus on the time window of 2026 is indeed increasing, but it is worth noting that—predictions of performance growth by a thousand or ten thousand times—this extrapolation framework is based on the assumption that AGI has been achieved, which carries significant uncertainty premium. The specific analysis is as follows: First, the current estimation methodology for the algorithm iteration space lacks supporting empirical data; second, the convergence of self-optimization remains controversial in academia. Risk warning—do not excessively extrapolate recent technological breakthroughs to directly anchor the 2030 forecast point.

View OriginalReply0

GasGuzzler

· 01-14 07:33

2026? Bro, if it really happens, we need to hop on the train quickly, or we'll be left behind. --- A hundredfold increase? That's a joke. It depends on whether the computing power keeps up, otherwise it's all just on paper. --- I just want to know if there will still be our share of work then... --- Free courses? Fine, since everyone is already competing fiercely, if you don't learn, you'll be overwhelmed anyway. --- If we really surpass humans by 2030, then nothing we learn now will matter, might as well just relax. --- Wait, why does it always feel like these predictions are moving forward... We also heard last year that a breakthrough would happen by 2025? --- Forget it, I'll just keep learning and watching. No one can predict the future anyway, but it's better than doing nothing.

View OriginalReply0