861–870 of 1001 posts
Earlier quoted context omitted.
-Chinese labs only have 5-10% the valuation of OpenAI/Anthropic Another reason could be that they want to close the gap between their valuations and OpenAI/Anthropic. Either Chinese labs are worth more or OpenAI/Anthropic are worth less. One of those is true
Does BYD want to close the gap between its valuation and Tesla's? I'd argue they're more concerned with capturing market share.
Earlier quoted context omitted.
Hey Simon, I noticed one thing all LLMs are currently pretty bad at and maybe we could create a benchmark from it. Let an LLM play the role of a dungeon master and tell it to strictly stay in the script/story and only allow realistic player actions. You will notice that they are easily brought off track. E.g. - Tell the LLM that you as a player noticed a strange glow in an NPCs eyes -> the NPC becomes an enemy. - In…
i found exactly the same thing trying to revive a text based game, a tool would be helpul.
I'm totally green when it comes to nlp,transformers, LLM training, etc, but I staunchly believe you can't produce real "reasoning" or consistent logic based on the predictions of byte pair encodings.
Earlier quoted context omitted.
If there was some grand strategy for all Chinese labs, surely it'd have leaked by now. I think its more likely that: - Companies can still make money from commodities - Chinese labs only have 5-10% the valuation of OpenAI/Anthropic, so massive monopoly profits aren't necessary. Profit expectations for tech companies in China are really low in general, complete opposite of the US. - Open weighting is a great way to ge…
-Chinese labs only have 5-10% the valuation of OpenAI/Anthropic Another reason could be that they want to close the gap between their valuations and OpenAI/Anthropic. Either Chinese labs are worth more or OpenAI/Anthropic are worth less. One of those is true
The reason that's not inherently accurate is because it assumes the economic models are identical. That couldn't be further from the truth. It'd be like pretending a globally dominate Google search engine should be worth as much as eg Baidu or Yandex. No doubt Baidu is an advanced search engine, and that's not what defines its economic characteristics: the markets it operates in does, the market ownership does, the ad engine does, the ability to sell does (advertiser relationships), etc.
OpenAI is very likely to bolt on a massive, global advertising platform to GPT over the coming years (and yes, it'll take time to build up). The Chinese labs will have a very hard time replicating that to a global audience the way the US tech companies have repeatedly proven they can. OpenAI will poach proven ad system builders from Meta and Google.
Earlier quoted context omitted.
Worry not, Pelicans on bicycles had been ranking pretty high on your favorite search engine for a while. I struggle to imagine a world in which it was not already scraped and turned into training data by at least one provider: 1. Models need to be good at the questions we ask them, not the questions we could ask them. 2. The questions, at least partially, are correlated with information people consume. 3. People most…
Probably it was added to the training data on the first day when this benchmark was on HN main page. It’s a bad benchmark since then. I don’t know why people still rate it high. Basically, every benchmark becomes pointless after it was published. They are good only to have a picture at the time they’re published first, and not after.
Earlier quoted context omitted.
I’m starting to come to the opposite approach: don’t try to customize anything, just use it vanilla, and use the best model you can afford. No AGENTS.md, no special subagents or roles, nothing but a few convenience skills which are really just textexpander. Use the harness that the LLM provider makes, and that’s it. Making a huge custom setup is so 2025.
I agree with this. However, aren't harnesses like Claude Code a bit bloated? Would it be better to use something like Pi?
Earlier quoted context omitted.
I had a thought a while back: sell large local models burned onto fused compute / ROM chips. Like cartridges for old game consoles. Slot (or probably plug into USB-C) and go. It’s an ASIC with the model wired into it so it’s very low power and fast. I’d buy these. Say $100 for a frontier class model. Maybe more.
isn't this essentially an FPGA?
> Chip Design > As an early proof of concept, Kimi K3 designed a chip to serve a nano model built on its own architecture. In a single 48-hour autonomous run, K3 built, optimized, and verified the chip using open-source EDA tools on the Nangate 45nm library. Within 4 mm², the chip closes timing at 100 MHz and sustains over 8,700 tokens/s decode throughput in simulation, packing 1.46M standard cells, 0.277 MB of SRAM,…