HNHacker News
TopNewBestAskShowJobs

fdsjgfklsfd

43 karma · joined September 4, 2020

submissionscomments
fdsjgfklsfd··on DeepSeek launching v4.1 flash cheaper and more capable than v4 pro
No you can't. It still responds in Chinese after explicitly asking it to "Always reason and respond in English."
fdsjgfklsfd··on DeepSeek-V4-Flash Update
Was already released before your comment: https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731
fdsjgfklsfd··on DeepSeek-V4-Flash Update
I use DeepSeek V4 Flash (before this update) for most things:

- OpenCode for codebase editing: python scientific computing and LLM projects

- Open Interpreter Classic (python version) for Swiss army knife terminal replacement one-off task type stuff.

fdsjgfklsfd··on DeepSeek-V4-Flash Update
DeepSeek-V4-Flash-0731 scores higher than Fable 5 on Terminal-Bench, and Fable 5 is Mythos, correct?
fdsjgfklsfd··on DeepSeek-V4-Flash Update
Yes, they trained on my open source code and Wikipedia edits and Stack Overflow answers that I shared freely, so it's only fair that they release their models as open source and share back to the community that created them. I only allow my training data go to open source models.
fdsjgfklsfd··on DeepSeek Introduces Vision
Yeah the reasoning is formatted differently and the replies are often in Chinese.
fdsjgfklsfd··on DeepSeek Introduces Vision
Sometimes it's faster than swyping on a phone, but mostly I use it to learn about stuff and hash out ideas while driving.
fdsjgfklsfd··on Qwen3.6-Plus: Towards real world agents
The US government is vastly more likely to break down my door and arrest me for my speech than the Chinese government is. (Because I live in the US.)
fdsjgfklsfd··on Qwen3.6-Plus: Towards real world agents
Qwen3.5-plus is quite good, Qwen3.6-plus is not.
fdsjgfklsfd··on Qwen3.6-Plus: Towards real world agents
Qwen3.5-plus has been my go-to model for non-autonomous chatbot that runs arbitrary code and shell commands on my local machine for one-off tasks (unlike Claude Code). It has no problem calling tools to run code (that contains tools), and pretty smart, and cheap (though lack of token caching is making it much more expensive than it should be).

I tried Qwen3.6-plus and it was not as good.

fdsjgfklsfd··on Qwen3.6-Plus: Towards real world agents
> In particular, Qwen3.5-Plus is the hosted version corresponding to Qwen3.5-397B-A17B with more production features, e.g., 1M context length by default, official built-in tools, and adaptive tool use. For more information, please refer to the User Guide.

https://huggingface.co/Qwen/Qwen3.5-397B-A17B

So 3.5-plus has been released as open weights.

fdsjgfklsfd··on Tell HN: Litellm 1.82.7 and 1.82.8 on PyPI are compromised
Reporting spam on GitHub requires you to click a link, specify the type of ticket, write a description of the problem, solve multiple CAPTCHAs of spinning animals, and press Submit. It's absurd.
fdsjgfklsfd··on We Will Not Be Divided
https://www.thestack.technology/backlash-over-anthropic-ai-c...

https://www.anthropic.com/news/detecting-and-preventing-dist...

fdsjgfklsfd··on We Will Not Be Divided
The current biggest problem in the US is that the President is violating the Constitution with impunity
fdsjgfklsfd··on We Will Not Be Divided
The Constitution and Founding Fathers are pretty great compared to what we have now.

"At this point, Elbridge Gerry objected to Butler’s earlier-raised proposition that the clause be shifted to a presidential power. Gerry remarked that he never expected to hear in a republic a motion to empower the Executive alone to declare war."

"What is called a republic is not any particular form of government. It is wholly characteristical of the purport, matter or object for which government ought to be instituted, and on which it is to be employed … in this sense it is naturally opposed to the word monarchy, which … means arbitrary power in an individual person; in the exercise of which, himself, and not the res-publica, is the object."

fdsjgfklsfd··on We Will Not Be Divided
Aggregators:

https://fiftyplusone.news/polls/approval/president

https://www.natesilver.net/p/trump-approval-ratings-nate-sil...

fdsjgfklsfd··on We Will Not Be Divided
"Overplayed"? Did you see the actual footage of the event? The event in which people attacked Capitol Police and broke into the Capitol and tried to take power by force?
fdsjgfklsfd··on Keep Android Open
They aren't actually trying to solve any real problem.
fdsjgfklsfd··on GPT-5
Do you mean "all variants of the same stacked transformer architecture converge in performance"? Or do you know of tests against some other architecture? The diffusion-based LLMs?
fdsjgfklsfd··on GPT-5
What's openai/gpt-5 vs openai/gpt-5-chat?
fdsjgfklsfd··on GPT-5
I think they're just reaching the limits of this architecture and when a new type is invented it will be a much bigger step.
fdsjgfklsfd··on Grok 4 Launch [video]
When I've had Grok evaluate images and dug into how it perceives them, it seemed to just have an image labeling model slapped onto the text input layer. I'm not sure it can really see anything at all, like "vision" models can.

It was giving coordinate bounding boxes and likelihood matches to generic classifications for each:

    - *Positions*:
      - Central cluster: At least five bugs, spread across the center of the image (e.g., x:200-400, y:150-300).
      - Additional bugs: Scattered around the edges, particularly near the top center (x:300-400, y:50-100) and bottom right (x:400-500, y:300-400).
    - *Labels and Confidence*:
      - Classified as "armored bug" or "enemy creature" with ~80% confidence, based on their insect-like shape, spikes, and clustering behavior typical of game enemies.
      - The striped pattern and size distinguish them from other entities, though my training data might not have an exact match for this specific creature design.
…

    - *Positions*:
      - One near the top center (x:350-400, y:50-100), near a bug.
      - Another in the bottom right (x:400-450, y:350-400), near another bug.
    - *Labels and Confidence*:
      - Classified as "spider" or "enemy minion" with ~75% confidence, due to their leg structure and body shape.
fdsjgfklsfd··on Grok 4 Launch [video]
I dunno. Talking with Grok 3 about political issues, it does seem to be pretty "truth-seeking" and not biased. I asked it to come up with matter-of-fact political issues and evaluate which side is more accurate, and it said the Left is more correct on almost all of them.
fdsjgfklsfd··on Grok 4 Launch [video]
Hello, LLM slop.
fdsjgfklsfd··on Grok 4 Launch [video]
You misspelled "principles".
fdsjgfklsfd··on Grok 4 Launch [video]
I feel like they should train a dumb model that does nothing but recognize when someone has finished talking, and use that to determine when to stop listening and start responding. Maybe it could even run on the phone?
fdsjgfklsfd··on Grok 4 Launch [video]
> you need to tap in order to end the response

I hope that can be turned off while driving...

fdsjgfklsfd··on Grok 4 Launch [video]
Cory Booker didn't do a Nazi salute. https://imgur.com/gallery/JwWQXSJ
fdsjgfklsfd··on Grok 4 Launch [video]
It confuses me that Elon is far-right in public, but names his creations from left-libertarian science fiction books. Is it just an act?
fdsjgfklsfd··on Grok 4 Launch [video]
Wait, are you implying that Grok 3 is "censored" because it aligns with "progressive" views?
Page 1 of 3Next →