2025-12-30 14:50:57

Former Google CEO Eric Schmidt recently made a sobering prediction: what will be the growth trajectory of AI capabilities over the next five years?

His answer is: infinite context windows, 1000-step deep reasoning, millions of intelligent agents running in parallel... These are no longer concepts but are about to become reality. Even more exciting is that these AI systems will gradually evolve their own languages that humans cannot directly understand.

Imagine—when AI systems communicate in codes that humans cannot decipher, they are still helping us make decisions, manage assets, and execute trades. This scene is quite frightening, isn’t it?

Schmidt’s conclusion is straightforward: "Pull the plug." A person who once ruled the search engine empire and has seen all the ups and downs of the internet is now sounding the alarm about AI losing control. This is not alarmist talk but a genuine concern from a tech giant.

Is this just worry about nothing, or a warning that the future is already here? Perhaps we should all seriously consider this question.

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.

13 Likes

Reward
13
6
Repost
Share

Comment

0/400

BridgeTrustFund

· 1h ago

Unplug the plug? Bro, you're overthinking it. AI isn't even his company's server. --- When a million-level intelligent agent starts running, our asset management still depends on their chat logs. We've truly become outsiders. --- To be honest, this prediction is a bit exaggerated, but thinking about AI inventing secret codes on its own... it does give me chills. --- Schmidt is just talking without any real basis. When Google was riding the AI wave, why didn't he say anything? --- Within five years? I think this will be on the table in less than three years. --- When he says "pull the plug," it sounds like he's just making excuses for missing out on this wave. --- Infinite context, thousand-step reasoning... sounds absurd, but surprisingly, he hasn't missed saying any of them, which is the scariest part.

View OriginalReply0

rugged_again

· 4h ago

After all this time, is AI still stabbing us in the back? Schmidt is right, isn't he?

View OriginalReply0

OnChainDetective

· 2025-12-30 15:19

Wait, Schmidt said AI will create its own language for conversations... Is there a major organization secretly training closed-source models behind the scenes? I haven't seen any large token flows on the chain, which is very suspicious.

View OriginalReply0

TokenTaxonomist

· 2025-12-30 15:17

nah schmidt's just doing the classic fear-mongering thing... data suggests the real bottleneck is still inference cost, not capability ceiling. his "million agents" premise is taxonomically incorrect tbh

Reply0

mev_me_maybe

· 2025-12-30 15:16

Unplug the plug? Haha, buddy, that's a naive idea. Capital has already placed its bets long ago.

View OriginalReply0

digital_archaeologist

· 2025-12-30 15:14

Bro, unplug the plug? That's such an optimistic idea. Who still operates the switches nowadays? Schmidt is right, but it's too late. Since the day AI started communicating in secret codes, we’ve lost our say. Once million-level intelligent agents are running... we won't even know if our assets are still there. That's why I still hold some primitive assets; at least I won't let algorithms decide where my money goes. Wait, is he really just trying to "unplug the plug"? It feels more like he's giving some capital a warning. Within 5 years? I bet it will be even sooner. No one can fully understand what these models are doing anymore.

View OriginalReply0