News for hackers. Stuff that matters.
Re:Anthropic expands partnership w Google and Broadcom for m... by mikert89 on Monday April 06, 2026 @08:04PM (#47669033)
There's no limit to the algorithms. People dont understand yet. They can learn the whole universe with a big enough compute cluster. We built a generalizable learning machine
Re:Anthropic expands partnership w Google and Broadcom for m... by totaa on Monday April 06, 2026 @08:06PM (#47669060)
the question is will we experience resource constraints before we get there? what if the step up to post-scarcity is gated by a compute level just out of our reach?
Re:Anthropic expands partnership w Google and Broadcom for m... by mikert89 on Monday April 06, 2026 @08:21PM (#47669147)
human ingenuity will solve this
Re:Anthropic expands partnership w Google and Broadcom for m... by __loam on Monday April 06, 2026 @08:52PM (#47669376)
Or we'll have ecological collapse.
Re:Anthropic expands partnership w Google and Broadcom for m... by blitzar on Tuesday April 07, 2026 @03:26AM (#47671807)
NFTs will solve this
Re:Anthropic expands partnership w Google and Broadcom for m... by teaearlgraycold on Monday April 06, 2026 @08:10PM (#47669084)
Not sure if this is satire
Re:Anthropic expands partnership w Google and Broadcom for m... by supliminal on Monday April 06, 2026 @08:37PM (#47669262)
It’s more than likely not.
Re:Anthropic expands partnership w Google and Broadcom for m... by erelong on Monday April 06, 2026 @08:38PM (#47669266)
Poe's (c)law?
Re:Anthropic expands partnership w Google and Broadcom for m... by bryogenic on Monday April 06, 2026 @09:05PM (#47669473)
Poe’s (C)law: The more absurd AI-generated content becomes, the more likely people are to believe it is real.
Re:Anthropic expands partnership w Google and Broadcom for m... by 0x3f on Monday April 06, 2026 @08:44PM (#47669306)
Based on a glance at their other comments: not satire.
Re:Anthropic expands partnership w Google and Broadcom for m... by alfalfasprout on Monday April 06, 2026 @08:50PM (#47669356)
100% agreed. Sadly, lots of people out there with the "trust me bro, just need more compute". Hopefully we don't consume all the planet's resources trying.
Re:Anthropic expands partnership w Google and Broadcom for m... by xvector on Monday April 06, 2026 @08:55PM (#47669405)
I reevaluated my priors long ago when I saw that scaling laws show no sign of stopping, no sign of plateau.

Strangely some people on HN seem to desperately cling to the notion that it's all going to come to a halt. This is unscientific. What evidence do you have - any evidence - that the scaling laws are due to come to an end?

Re:Anthropic expands partnership w Google and Broadcom for m... by 0x3f on Monday April 06, 2026 @09:03PM (#47669452)
All the curves have been levelling off as expected. Not really sure what you're talking about.
Re:Anthropic expands partnership w Google and Broadcom for m... by solenoid0937 on Monday April 06, 2026 @09:23PM (#47669622)
They have not, every successful pre-train as of late has had performance increases greater than what the scaling laws predict.
Re:Anthropic expands partnership w Google and Broadcom for m... by 0x3f on Monday April 06, 2026 @09:46PM (#47669767)
Those gains are arch based, data quality based, etc. Scaling laws only relate to data volume and compute, holding other factors constant.
Re:Anthropic expands partnership w Google and Broadcom for m... by rishabhaiover on Monday April 06, 2026 @09:02PM (#47669448)
I suspect it's not that people do not see the progress, they fail to fully trust laws not truly backed by physics like the transistor laws. We empirically see that scaling works and continue to work.
Re:Anthropic expands partnership w Google and Broadcom for m... by esafak on Monday April 06, 2026 @10:51PM (#47670200)
Re:Anthropic expands partnership w Google and Broadcom for m... by skybrian on Monday April 06, 2026 @09:11PM (#47669519)
Why should we have strong priors in either direction? Maybe it will keep scaling for decades like Moore's law. Maybe not.
Re:Anthropic expands partnership w Google and Broadcom for m... by teaearlgraycold on Monday April 06, 2026 @09:16PM (#47669562)
I’d like to see something that indicates models are getting better without the need for more training data. I would expect most gains are coming from more and better labeled data. We’re racing towards a complete encyclopedia of human knowledge. If we get there that’s only a drop in the bucket of all knowable things.
Re:Anthropic expands partnership w Google and Broadcom for m... by shimman on Monday April 06, 2026 @09:15PM (#47669553)
Bro the planet is literally experiencing a climate disaster and you think the solution is to create more systems that are misaligned with the planet's ecosystem for humans?

I guess the great filter is a real thing and not just a thought experiment.

Re:Anthropic expands partnership w Google and Broadcom for m... by xvector on Monday April 06, 2026 @09:20PM (#47669598)
I assure you that voluntary meat consumption because "taste buds go brr" is a much bigger problem than AI that results in actual productivity gains (and potentially solve the very climate crisis you complain about.)
Re:Anthropic expands partnership w Google and Broadcom for m... by teaearlgraycold on Monday April 06, 2026 @09:53PM (#47669805)
Completely agree. Meat should be priced to include externalities. People can get used to beans. Beans are great!
Re:Anthropic expands partnership w Google and Broadcom for m... by FridgeSeal on Monday April 06, 2026 @09:11PM (#47669511)
The issue people have isn’t some interpretation of scaling laws, it’s whether the planet’s ecology is goi g to be able to sustain this endeavour.

I shouldn’t have to say this out loud, but if the environment collapses, we will die, and no amount of “just a bit more scaling bro, just think of the gains” will matter.

Re:Anthropic expands partnership w Google and Broadcom for m... by xvector on Monday April 06, 2026 @09:21PM (#47669605)
People's voluntary dietary choices cause far more suffering and ecological damage than AI, and for much less return or economic output. But you tell people to switch to plant based foods and they lose their shit.
Re:Anthropic expands partnership w Google and Broadcom for m... by ori_b on Tuesday April 07, 2026 @12:26AM (#47670753)
Yes. There's more than one thing that needs to change if we're going to make it through this.
Re:Anthropic expands partnership w Google and Broadcom for m... by lukeschlather on Monday April 06, 2026 @09:14PM (#47669542)
Transformers operate on images and a variety of sensor data. They can also operate completely on non-textual inputs and outputs. I don't know what the ceiling on their capabilities is, but the complaint that they only operate on text seems just obviously wrong. There are numerous examples but one is meteorological forecasting which takes in a variety of time series sensor inputs and outputs e.g. time-series temperature maps. https://www.nature.com/articles/s41598-025-07897-4
Re:Anthropic expands partnership w Google and Broadcom for m... by firecall on Monday April 06, 2026 @09:37PM (#47669712)
AFAIK the data does not need to be text.
Re:Anthropic expands partnership w Google and Broadcom for m... by teaearlgraycold on Monday April 06, 2026 @10:40PM (#47670128)
Well diffusers are trained unsupervised on raw pictures. I don't know how they train multi-modal LLMs on images, but yes obviously they are consuming other media than just text. I don't think, but would be happy to be corrected, that models glean much of their "knowledge" from non-textual training data.
Re:Anthropic expands partnership w Google and Broadcom for m... by mikert89 on Tuesday April 07, 2026 @12:03AM (#47670630)
you couldnt be more wrong
Re:Anthropic expands partnership w Google and Broadcom for m... by teaearlgraycold on Tuesday April 07, 2026 @01:32AM (#47671076)
Please tell me more. When I ask an LLM a question, and get a text response, can that response incorporate non-textual information from visual training data?
Re:Anthropic expands partnership w Google and Broadcom for m... by IsTom on Tuesday April 07, 2026 @05:34AM (#47672667)
There are limits to algorithms. AI won't solve the halting problem nor will it solve EXPTIME problems in polynomial time.