Palmik

7 ms·

91.
by Palmik 10mo ago
I think possibly even better would be viral, GPL-like license that explicitly mandates that any systems (models, etc.) derived (trained on) the code need to be released under the same license.
92.
by Palmik 10mo ago
It would not be discrimination to mandate that weights of any model trained in the code need to be released under similarly open license.
93.
by Palmik 10mo ago
Hard disagree. At Google, in most orgs, manager can influence the chance of success significantly: - Making sure their team works on what the org leads find "impactful" - Facilitating cross team collaborations, which will lead to
94.
by Palmik 10mo ago
There are plenty of NSFW oriented apps, especially in the AI category. > It's popular with women and basically invisible to men. Mostly true, and this might be a reflection of reality, but certainly not a justification.
95.
by Palmik 10mo ago
I wonder why Texas did not start by targeting NSFW / porn apps specifically, like other states. I also wonder why smut literature (the best selling category of books on Amazon) seems to get a free pass.
96.
by Palmik 10mo ago
Children will always be able to use devices or accounts borrowed or bought from adults, regardless of how the initial verification is carried out. Not to mention that the verification key / token / device might also be borrowed or
97.
by Palmik 10mo ago
I think calling Open AI Codex or Claude Code "CLI" is a bit a of minomer. It's more of a GUI, just rendered in a terminal. I honestly think a "regular" for GUI for OpenAI Codex / Claude Code could be much bette
98.
by Palmik 10mo ago
In context of AI I'm mostly seeing anti-OpenAI pro-Google bias.
99.
by Palmik 10mo ago
Why would you use OpenRouter rather than some local proxy like LiteLLM? I don't see the point of sharing data with more third parties and paying for the privilege. Not to mention that for coding, it's usually more cost efficient t
100.
by Palmik 10mo ago
No offense, but that seems like a poor benchmark. These initial vibe checks are easily swayed by personal brand biases.
101.
Denmark wants to ban VPNs to unlock foreign, illegal streams(techradar.com)
8 points by Palmik 10mo ago|3 comments
102.
ChatterBox Turbo – Open Weight TTS Model(huggingface.co)
2 points by Palmik 10mo ago|0 comments
103.
by Palmik 10mo ago
Jailed for what? Apple routinely buys out majority of TSMC's production capacity (at the bleeding edge), should Tím Cook go to jail?
104.
by Palmik 10mo ago
If age verification has to be a thing (which isn't necessarily clear), then it should be up to the client to provide age indication to the server. It could be through a header, or something like this: https://developer.chrom
105.
by Palmik 10mo ago
That does not make the title any less clickbaity. Moreover, it does not seem like a vindication of johnfn's original comment.
106.
by Palmik 10mo ago
DeepSeek and other Chinese companies. Not only do they publish research, they also put their resources where their mouth (research) is. They actually use it and prove it through their open models. Most research coming out of big US labs is
107.
by Palmik 10mo ago
DeepSeek's MLA paper was published in 2024: https://arxiv.org/abs/2405.04434 DeepSeek's Sparse Attention paper was published in February: https://arxiv.org/abs/2502.11089 DeepSeek 3.2 Ex
108.
by Palmik 10mo ago
This is second time this week: https://news.ycombinator.com/item?id=46140145 The previous one affected European users for >1h and made many Cloudflare websites nearly unusable for them.
109.
by Palmik 10mo ago
The status pay is egregiously misleading, it definitely affects more than just their dashboard.
110.
Ask HN: Is Cloudflare Down for You?
20 points by Palmik 10mo ago|12 comments
111.
by Palmik 11mo ago
It's a fair question, even if it might be coming from a place of misunderstanding. For example, DeepSeek 3.2, which employs sparse attention [1], is not only faster with long context than normal 3.1, but also seems to be better (perhap
112.
by Palmik 11mo ago
GPT 5.1 / Codex already beats Gemini 3 on SWE Bench Verified and Terminal Bench and this pushes the gap further. Seems like a decent improvement.
113.
by Palmik 11mo ago
All evals on Terminal Bench require some harness. :) Or "Agent", as Terminal Bench calls it. Presumably the Gemini 3 are using Gemini CLI. What do you mean by "standard eval harness"?
114.
by Palmik 11mo ago
All evals on Terminal Bench require some harness. :) Or "Agent", as Terminal Bench calls it. Presumably the Gemini 3 are using Gemini CLI.
115.
by Palmik 11mo ago
The reported results where GPT 5.1 beats Gemini 3 are on SWE Bench Verified, and GPT 5.1 Codex also beats Gemini 3 on Terminal Bench.
116.
by Palmik 11mo ago
Also does not beat GPT-5.1 Codex on terminal bench (57.8% vs 54.2%): https://www.tbench.ai/ I did not bother verifying the other claims.
117.
by Palmik 11mo ago
GPT 5 was released more than 3 months ago. Gemini 2.5 was released less than 8 months ago.
118.
by Palmik 11mo ago
Archive link: https://web.archive.org/web/20251118111103/https://storage.g...
119.
by Palmik 11mo ago
Then surely you would not be opposed to Rebble using copy-left OSI compatible license, right?
120.
by Palmik 11mo ago
> To both implement an age verification system and also to block the access of users connected via VPN. This would require blocking VPN use for all users, regardless of where the user comes from (at just until their age verified). That s

More ›