esafak

21 ms·

1. 661.▲
   by esafak 5mo ago

   Good on Nadella: After expressing his complete disbelief that such a document could have been written, Nadella adds that the elusive and mysterious authors “may want to go work elsewhere.” edit: VP of a product I had not even heard of; it
2. 662.▲
   by esafak 5mo ago

   Because it could discard things the agent needs.
3. 663.▲
   by esafak 5mo ago

   I suggest making your submission title more informative. Can you talk about the motivation for rewriting TigerBeetle?
4. 664.▲
   by esafak 5mo ago

   I have no idea what this does and I'm not inclined to find out. A short video and screenshots would help.
5. 665.▲
   by esafak 5mo ago

   We'd need a benchmark to tell.
6. 666.▲
   by esafak 5mo ago

   Ephemeral environments
7. 667.▲
   by esafak 5mo ago

   Coloring books give you some freedom; to pick your own colors and texture. If we were to draw an analogy, it would be using the LEGO pieces in this kit to build something else. But nobody is going to do that, so it is like a color-by-number
8. 668.▲
   by esafak 5mo ago

   > In the coming months, we will organize conversations where policymakers, researchers, civil society, and other AI companies can help answer some of the questions this piece raises, especially around full recursive self-improvement and
9. 669.▲
   by esafak 5mo ago

   It's the output of a research paper; the authors are not trying to build up vLLM, and they probably have no incentive to do so. You can submit a PR, though! It's easier now while the divergence is low, so don't wait. Since th
10. 670.▲
    by esafak 5mo ago

    What's your job title?
11. 671.▲
    by esafak 5mo ago

    https://en.wikipedia.org/wiki/NATS_Messaging is for real-time communication, so I would not use it as a substitute for a memory store.
12. 672.▲
    by esafak 5mo ago

    As mentioned, Vite is already Rust, plus the same developers (the subject of the news) have developed https://viteplus.dev/
13. 673.▲
    by esafak 5mo ago

    That's why I always ask: got benchmarks?
14. 674.▲
    by esafak 5mo ago

    I made a mistake; apparently Glean uses knowledge graphs too: https://www.glean.com/resources/guides/glean-knowledge-graph This raises a follow-up question: what is your differentiation?
15. 675.▲
    by esafak 5mo ago

    1. Have you measured the value provided by the knowledge graph layer over straight enterprise search (e.g., https://www.glean.com/ ) Benchmarks, please. 2. How do you deal with conflicting facts? In tech, the new is constant
16. 676.▲
    by esafak 5mo ago

    Way more. Track with https://github.com/junhoyeo/tokscale
17. 677.▲
    by esafak 5mo ago

    Since electricity demand has outpaced supply they are increasing prices: https://www.bloomberg.com/graphics/2025-ai-data-centers-elec...
18. 678.▲
    by esafak 5mo ago

    Coming, luser. Moove! -- Nick Burns, the computer guy.
19. 679.▲
    by esafak 5mo ago

    Could you publish terminal-bench scores? What about memory usage?
20. 680.▲
    by esafak 5mo ago

    What can you say about it?
21. 681.▲
    by esafak 5mo ago

    I don't understand why you wouldn't simply provide SQL with requisite security precautions. Do you really need more?
22. 682.▲
    by esafak 5mo ago

    Is there a need for this over simply providing a rich enough search API?
23. 683.▲
    by esafak 5mo ago

    If anything text lovers prefer GUIs for their proper rendering. Imagine if books looked like TUIs.
24. 684.▲
    by esafak 5mo ago

    I'm pleased to see SurrealDB making progress. I look forward to the distributed comparisons. Can you also report TPC benchmarks? Also, do you have any write ups on your correctness- and regression testing approach?
25. 685.▲
    by esafak 5mo ago

    orbstack and colima work well
26. 686.▲
    by esafak 5mo ago

    The screenshot is illegible. Please compare it to competing products in the README. Instructions on adding support for new languages in CONTRIBUTING.md would help too.
27. 687.▲
    by esafak 5mo ago

    I suppose it depends on what kind of users you have in mind; enthusiasts, versus average users. Before they became outright user-hostile they were known for their anti-competitive behavior and buggy products. People were calling them "
28. 688.▲
    by esafak 5mo ago

    Again, this analogy makes no sense. People use PowerPoint, they don't get replaced by it. The goal of AI companies is to replace workers entirely; that is the only way their valuations make sense, and OpenAI's charter says this ex
29. 689.▲
    by esafak 5mo ago

    How does that analogy apply to AI, where a handful of companies are attempting to replace the entire white collar market with computers? It fits neither qualitatively, nor quantitatively.
30. 690.▲
    by esafak 5mo ago

    Software engineering is flattening. I wouldn't go into 'pure' app development but specialize in a regulated field, unless I had entrepreneurship aspirations, in which case now is a great time to enter the field.