Recent submissions

Comment · 5 days ago
Everything old is new again - return of the phablet.
Comment · 5 days ago
I don't participate in stock markets for personal reasons but if I did, I would be betting against Anthropic in the long term. This is a highly commoditized field and while they are a leader it's not clear than any of their attempts at vendor lock-in are working. They are certainly not working on me as I cancel and enable my subscription depending on the availability of capable lower-cost models.
Comment · 6 days ago
I thought the price reduction was because both companies implemented Deepseek's published research which demonstrate effective approaches to lowering the cost of inference while maintaining quality. Thanks to Deepseek, they did lower their costs and they passed that in to their customers.
Comment · 6 days ago
My kids must be losers - they still like books.
Comment · 9 days ago
Yes, Jared Kushner is a human. A human criminal.
Comment · 14 days ago
The world turns a blind eye to Israeli terrorism.
Comment · 18 days ago
I always wanted an immovable laptop that is closer to phone size. I'm just a worst of both worlds kids of guy.
Comment · 26 days ago
This is like an ad for how great Deepseek V4.1 Flash is.
Comment · 28 days ago
Kushner is like the human embodiment of conflict of interest. I can't see his face without immediately thinking the term: corruption.

He's like the fresh face of elite capture.

Comment · 28 days ago
Yes, this makes sense. Memory is an uncurated and often opaque system of arbitrary past discussions. It can help, it can harm. Accurate documentation in the other hand is only beneficial.
Comment · 28 days ago
For the benchmark... it seems a bit silly to compare only to outdated/underperforming models.
Comment · 1 month ago
People's volume and approach varies. I'm a happy customer and I use my entire double max subscription on planning and analysis and have other models doing all my implementation work because I would burn through my subscription in a day or less. It's difficult to calculate, but I'm something like 10-20 billion token per week consumer and I can't use a US-based model to do this volume of implementation work.

Also, a lot of this work is verification to ensure that AI generated code does what is intended and is safe to merge and deploy. That verification work is critical and uses a lot of tokens.

Comment · 1 month ago
Well articulated. Ty
Comment · 1 month ago
Well, I don't go to website anymore... so ha! ;)
Comment · 1 month ago
It can be both. They were already doing it and the situation made it more urgent.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton