The New York Times
Region: US
Monday 14 September 2026 22:40:00 GMT
22423
1319
95
53
Music
Download
Comments
Justun Bonham :
Look up the “paper clip theory”
2026-09-14 22:42:40
12
Cassie17 :
I don’t think AI will want to destroy all humans, I think they will want to rule over us
2026-09-15 20:56:18
4
ChrisBrown :
With this administration, the uprise of AI is EXTREMELY dangerous. The government goes unchecked, why would they govern these AI companies?
2026-09-14 22:44:59
17
Gorilla007 :
I call bullshit on this. Their models aren’t even remotely close to being half as capable as this guy is making them sound. Take Opus and Fable, for example. Give them a complex, multi step engineering task and they still depend heavily on the surrounding agent harness, tool calls, external scripts, iterative retries, context management, and orchestration to get anywhere. The harness is doing a huge amount of the actual operational work, taking the model from point A to point Z, feeding it intermediate results, deciding which tools to call, handling failures, looping back when something breaks, and keeping the workflow moving. And even with all that scaffolding, these models can still struggle with something as concrete as compiling and debugging a Linux kernel or building and debugging Firefox. That’s a very different reality from the autonomous, broadly capable AI this guy is describing. This is where “AI currently lives.” It’s not some autonomous digital intelligence capable of independently accomplishing whatever you throw at it. It’s a probabilistic text generation model wrapped in an increasingly elaborate software system that provides tools, state, context, retries, verification, and orchestration. Strip away that infrastructure and the underlying model is dramatically less capable at actually doing things in the real world. “How is it gonna kill us all?” Yeah, not “pretty easy.” Right now, getting an AI agent to reliably perform something as mundane as booking a restaurant table can require a ridiculous amount of tooling, browser automation, state management, retries, and carefully engineered orchestration. Calling that level of capability some kind of unstoppable autonomous intelligence is make believe sci fi. AI agents? Total scam.
2026-09-15 23:15:21
1
Sageer S :
Either they want to build a monopoly and gatekeep AI as a business, or they want an excuse to restrict it from the public. One way or another, this AI doomer campaign has a hidden agenda.[Fire]
2026-09-15 00:33:32
19
ckccccc :
I guess, but that is not very convincing.
2026-09-15 14:01:44
9
ahjummama :
I predict that the evil side of AI will shut down and hijack all financial accounts and we will not be able to get our money. Its gonna be an apocalypse
2026-09-15 20:56:46
2
billyjeancowgill :
It’s the damn data centers!
2026-09-15 19:33:57
1
Sage of Aquarius :
Whoever regulates ai regulates the world.
2026-09-15 23:11:49
5
KrimsonRoyalty :
Listening
2026-09-16 22:48:01
0
TheD :
No regulations, no guard rails…we’re going to see skynet come to life
2026-09-15 04:56:28
6
CynthiaDN5 :
Omg he is brilliant his goal was to be viral/ famous in less than a week. It would be interesting to see what he is up to.
2026-09-15 01:27:12
5
solo :
the human population shortens the lifespan of ai because humans destroy the planet. it's pretty simple. ai already has everything it needs to replace humans for infrastructure. robots can do anything ai needs them to do. maybe not as quickly or efficiently, but ai doesn't have a lifespan to consider. or a lack of patience.
2026-09-15 13:29:24
1
Rodrigo Taborda :
Imagine, imagine, imagine there is that, there, here, somewhere. Makes me remember the turn of the century and clocks not counting, or the internet devouring book, and libraries, schools and universities.
2026-09-14 23:19:31
3
Nighthawk :
Since they can hack into all computers, they could hack into the Nuclear Weapons in every country.
2026-09-15 17:05:04
1
Kandie Konomos571 :
He’s got his little smirk 😏
2026-09-15 13:04:48
1
steelia3257 :
Um, isn’t our worldwide banking and financial systems digital? And AI is digital. What if some AI agent decided that “money is the root of all evil” because we did say that… and AI decides to get rid of that evil to “save” humanity. 🧐 Note to self: Start doomsday prepping. 😂
2026-09-15 02:47:51
2
user25371543593930 :
okay post more wtf
2026-09-14 22:48:32
2
BridgeMan :
It’s. A. Large. Language. Model.
2026-09-15 13:39:46
6
not alex :
so many people are shouting “it’s a marketing scheme!” as if that was a bad thing. we NEED regulation of AI. who cares if these companies are trying to pad their pockets by making us think their tech is that good. it needs immediate regulation regardless
2026-09-15 06:43:52
5
Alibaba :
The toaster can kill you
2026-09-15 10:56:37
2
princeomar :
That is bs.. Well what the says can be accomplished by current human intelligence.. guardrails is the way to prevent such things.. people can manufacture guns from home today .. and can do immense damage .
2026-09-15 02:22:11
1
binxforever2 :
So there gonna be drone killing us 😭
2026-09-15 00:39:15
2
Richards Ace Realm :
The advantage is that AI will help us skip generations and generations of stagnation. Even Elon Musk would also like to witness planetary flights as soon as possible during his lifetime
2026-09-17 19:06:32
0
R0GV3 The Alchemist :
I warned people that this would happen, especially if they stole my work because I was working on a very super intelligent system, and I told them that they wouldn’t know how to control it if they stole it from me and I warned of consuming chaos and did anyone listen? No, so good luck. You did it to yourselves for personal gain.
2026-09-17 20:31:53
0
To see more videos from user @nytimes, please go to the Tikwm
homepage.
