dakolli

7 ms·

1. 151.▲
   by dakolli 3mo ago

   They don't even beat 3.5 flash.. I'm really not sure where you're getting this from.. vibes? 3.5 flash beats out even fable on tool calling, which is really all that matters.
2. 152.▲
   by dakolli 3mo ago

   What are you talking about gemini 3.5 flash beats fable at tool calling and is 5x faster... I think it's very competitive and what most normal people are using.
3. 153.▲
   by dakolli 3mo ago

   I won't use anything created by elon. Who cares.
4. 154.▲
   by dakolli 3mo ago

   I mixed up some names. The timing doesn't seem coincidental. We are at the end of Executive Order 14301, signed May 2025, which called for at least three test reactors to reach criticality by July 4, 2026. So immediately after Trumps
5. 155.▲
   by dakolli 3mo ago

   because they have infiltrated the government to reduce the cost of safety, and increase the possibility of environmental harm to pad their margins... faster shit code, AI cat videos and so they can add 100ft to the length of their next boa
6. 156.▲
   by dakolli 3mo ago

   A bunch of these nuclear power startups have started reached criticality over the last week. Aalo and Valar (thiel) and now GAO is trying to loosen regulations around nuclear waste disposal.. Makes sense. Weird how we only get green energy
7. 157.▲
   by dakolli 3mo ago

   because they don't know what they're doing.
8. 158.▲
   by dakolli 3mo ago

   there's huge margins on GPU time, not tokens.
9. 159.▲
   by dakolli 4mo ago

   people that use vim motions/shortcuts/keyboard workflows are more productive, this is undeniable..
10. 160.▲
    by dakolli 4mo ago

    I have found llms to be utterly useless for frontend (tailwind included). That is, unless you're building a single page app/landing page that is the typical center column with a hero and below that a 3x3 feature grid with those sa
11. 161.▲
    by dakolli 4mo ago

    Why would you believe anything an intelligence agency says? Half their job is spreading fear and obfuscating the truth.
12. 162.▲
    by dakolli 4mo ago

    All the code an LLM produces is of questionable quality, so I'm not sure why you'd prioritize quality over speed. Speed is their only value add.
13. 163.▲
    by dakolli 4mo ago

    Nobody who has ambitions of being the top of their field in engineering wants to be a water treatment specialists, arborist or actuary (maybe actuary if you're a stats nerd). What you're saying is go do something on the based on t
14. 164.▲
    by dakolli 4mo ago

    While Trump is terrible, all the same morally questionable practices existed under Clinton, Bush, Obama, Biden. This administration just likes to brag about it. The US has been controlled by an evil technocracy/intelligence apparatus f
15. 165.▲
    by dakolli 4mo ago

    There's no way to safely use SOTA LLMs if privacy, and IP protection are your concern. Unless you want to spend 100k+ to host a 1T param model. Even if you use OpenCode you're sending all that information to random data centers yo
16. 166.▲
    by dakolli 4mo ago

    for using a proxy service.. wtf
17. 167.▲
    by dakolli 4mo ago

    Are you living in a cave? The US just sent someone to prison for 30 years for passing out political magazines at a protest. https://www.theguardian.com/us-news/ng-interactive/2026/jun/... The US has a la
18. 168.▲
    by dakolli 4mo ago

    Most people don't care about trust anymore, we live in a low trust society where this is to be expected. People gladly line up to be poisoned by fast food restaurants and trade 1/3 of their life for pieces of paper on a daily basi
19. 169.▲
    by dakolli 4mo ago

    I trust Chinese companies with my data way more than the corporations of the 4th Reich.
20. 170.▲
    by dakolli 4mo ago

    How to light 6k on fire.
21. 171.▲
    by dakolli 4mo ago

    If this is the most important industry for the US economy, the US is screwed. and .... Hopefully you're correct. In past Empires kings bet their entire nations future on the words of soothsayers , people who said they could predict the
22. 172.▲
    by dakolli 4mo ago

    https://aimultiple.com/gpu-benchmark concurrency
23. 173.▲
    by dakolli 4mo ago

    I'm being very sarcastic, local model evangalists seems to just be operating on vibes when they say these things and are completely disconnected from how models work, what the hardware requirements are. Prices aren't going down, a
24. 174.▲
    by dakolli 4mo ago

    That too.
25. 175.▲
    by dakolli 4mo ago

    Sure, If you want to light money on fire for entertainment, more power to you. There's probably worse ways to light 100k on fire. If I have an extra 100k laying around it's going to my family though.
26. 176.▲
    by dakolli 4mo ago

    Cool then everyone will just change their config to route through a provider overseas for an added 50-100ms latency. Who cares.
27. 177.▲
    by dakolli 4mo ago

    8 X RTX6000. It will run you around 80-100k to get started with a model at this size with decent tps.. Don't worry though, open source evangelists will tell you that these will be running on your phone in the next 3 years. For $100k y
28. 178.▲
    by dakolli 4mo ago

    Weird https://ipcrawl.com/?page=6&cam=63f7feaf5042d223
29. 179.▲
    by dakolli 4mo ago

    Its because our culture worships pieces of paper the government tells us is worth something.
30. 180.▲
    by dakolli 4mo ago

    Okay dude, whatever makes you feel good. I don't want to interrupt your cope, you're right, China can't compete with the West.