News for hackers. Stuff that matters.

hmate9

Re:OpenAI Withdraws 3 Math Papers on Thursday October 08, 2026 @07:53AM (#50004733)
3 mistakes (so far) out of ~400 is still a pretty good hit rate
Re:On the Navier–Stokes Millennium Prize Problem on Tuesday September 08, 2026 @01:22PM (#49613410)
Napkin math if we assume gpt 6 astra on max is >$15 million (just for output tokens) for those wondering.
Re:Gemini 3.8 Flash on Wednesday September 02, 2026 @12:15PM (#49538498)
It is more expensive per task than 5.6-sol high: https://artificialanalysis.ai/models/gemini-3-8-flash#price-...

Dropbox Data Breach

I received a security notice from Dropbox today saying that my account was accessed without authorization between August 4 and August 21, 2026, and that Dropbox believes files in the account were viewed or downloaded.According to the email, Dropbox uses Lenovo as an identity provider, allowing users to authenticate to Dropbox with a verified Lenovo ID.Dropbox says:an issue with Lenovo’s email...

( Read More... | 8 comments | 17 points | on HN )
Re:Gemini 3.6 Flash on Tuesday July 21, 2026 @11:24AM (#48993514)
This model is not for builders and engineers. DeepSWE score of 49% is behind gpt 5.4 and muse spark. It's clearly intended to be an efficient model for google gemini usage.

What is interesting is how this is announced before any Gemini Pro progress. From the outside it seems as though Google cannot keep up with other frontier models.

Re:Show HN: One More Letter on Thursday July 16, 2026 @03:59AM (#48931557)
Thank you guys for the feedback. I'll fix valid words being rejected, having space as a keyboard shortcut for shuffle, and maybe a "zen" mode without a timer.
Re:Previewing GPT‑5.6 Sol: a next-generation model on Friday June 26, 2026 @01:15PM (#48689130)
What is this prediction based on?
You can pay the most if you can get the most value out of it
Not defending this, and it's far from ideal, but also credit card details were already pretty much confirming identities already.
This is incredibly bullish for china and open source models
Re:DiffusionGemma: 4x Faster Text Generation on Wednesday June 10, 2026 @01:05PM (#48479333)
I can’t help but feel like there’s something here that will matter for future LLMs.

The bidirectionality could be a big deal: being able to refine a sentence with both left and right context feels closer to how editing/thinking actually works than committing to each token forever.

Maybe the current models aren’t good enough yet, but the direction feels important.

Re:Gemini 3.5 Flash on Tuesday May 19, 2026 @04:49PM (#48199431)
I have google ai pro plan and tried antigravity with 3.5 flash but it used up all my quota in two prompts. If that is not a bug then it is seriously unusable.
Re:Updates to GitHub Copilot interaction data usage policy on Wednesday March 25, 2026 @04:39PM (#47522894)
going to github i was greeted with a banner and a link directly to the settings for changing it
Re:Updates to GitHub Copilot interaction data usage policy on Wednesday March 25, 2026 @03:46PM (#47522239)
For what it's worth they're not trying to hide this change at all and are very upfront about it and made it quite simple to opt out.
I think the burden to curate your feed so that you do not have such content is now resting with the user and they cannot rely on the platform to do it for them.
Is this unavoidable? I mean it does generate clicks and views and user engagement so if one platform is doing it, doesn't that automatically mean that the other has to do it? Otherwise they will continuously lose market share.

Ask HN: Are past LLM models getting dumber?

I’m curious whether others have observed this or if it’s just perception or confirmation bias on my part. I’ve seen discussion on X suggesting that older models (e.g., Claude 4.5) appear to degrade over time — possibly due to increased quantization, throttling, or other inference-cost optimizations after newer models are released. Is there any concrete evidence of this happening, or technical...

( Read More... | 5 comments | 3 points | on HN )
Re:Testing Ads in ChatGPT on Monday February 09, 2026 @02:29PM (#46949811)
I hate ads too but whats the outrage? Did people expect it to be free forever? Everything else has ads. youtube, instagram, x, google etc. etc.
Re:xAI joins SpaceX on Monday February 02, 2026 @04:58PM (#46862293)
Any details regarding valuations etc?
Re:The Codex App on Monday February 02, 2026 @02:50PM (#46860421)
That’s not really a negative for me as I can easily jump into vscode where I already have my workspace for coding set up exactly as I like it. This being a completely separate app just to get the agentic work right is a good direction imo
About 600GB needed for weights alone, so on AWS you need an p5.48xlarge (8× H100) which costs $55/hour.
Re:Venezuela's capital hit by explosions, smoke seen on Saturday January 03, 2026 @11:57AM (#46478855)
As expected: "Trump now turns now to oil. He claims the oil business in Venezuela has been a 'bust', and that large US companies are going to go into the country to fix the infrastructure and "start making money for the country"."
Re:Venezuela's capital hit by explosions, smoke seen on Saturday January 03, 2026 @10:09AM (#46477535)
"Venezuela has ~2.7× as much proved oil as the UAE (303 / 111 ≈ 2.73)."

Pretty incredible

i understand im just saying that the other 99% of the time this signal leads to nowhere so i would not pay attention to it
There are like 10 of these surge pizza traffic posts from this account every month. And 99% of the time we don’t hear anything happening afterwards so I’m not sure how reliable it is
Re:Nvidia buying AI chip startup Groq for about $20B on Wednesday December 24, 2025 @06:07PM (#46380278)
There is also movement over of some key people to nvidia which is pretty significant: "As part of this agreement, Jonathan Ross, Groq’s Founder, Sunny Madra, Groq’s President, and other members of the Groq team will join Nvidia to help advance and scale the licensed technology."

Older »