Hacker News Reader

Comment by culi | Hacker News Reader

A third the cost, open source, and won't refuse every other request because of some vague possible connection to cybersecurity concerns.

View on HN

There is too much business risk in running on an American company. They can pull the model back or lobotimize it. No Alex Karp fan, but he was right: companies are worried about hyperscalers stealing their alpha. The lack of guardrails, data security and control really give these open models the edge. If you factor in that they appear to cost less, the hyperscalers are in big trouble. I don't see how this works out. Software was always supposed to be deflationary and collapse down to 0 marginal cost but this isn't remotely the case. It's truly amazing to see the state of open weight models

My questions about strawberries got blocked as too dangerous! I’m not joking

Strawberries are a well known weakness of LLMs, as they have a hard time to count the numbers of "r"s in them. Maybe that's why, because they fear that weakness could be exploited somehow.

I always thought this should be easier by telling it to write each letter one a new line, then count, any token separator ought to suffice, so much so you'd think they'd have trained in this strategy given tokens make individual letters opaque

Strawberries aren't the weakness, the weakness is the tokenization of a prompt. Any word with multiple duplicate characters is going to be troublesome for LLMs.

I think the parent is aware and was joking.

Probably need to be taught by someone of Latino origin. Learning to roll them "r"s could help.

2/3rds of all languages use rolled r's

2/3rrrrds of all languages use rolled r's

Really? I find that hard to believe and can't find a reference for it.

My God! You're practically a terrorist and should be on a watchlist! Who knows what you'll ask next? "Who is the surgeon to the boy?"

I asked about Tiananmen Square and it said it knew nothing about it. A real life “Doesn’t look like anything to me” moment.

Or biology or chemistry.

Or weapons manufacturing. It can refuse to answer on a lot of topics.

From the blog post moonshot refers to it as open source but only mentions releasing the weights.

So far all the Kimi models have open sourced their code, their weights, and published technical reports explaining their training methodology. I expect the Kimi K3 Technical Report will come out July 27 and they usually publish the code and the weights alongside that.

The weights are the "source" of a model.

If weights are the source for models then ELF binaries are the source for software.

Everything is open source if you know how to reverse engineer ;)

Then everyone is open source if you have $20 to spend on tokens, I guess.

clearly not true. the weights are the preferred form for making modifications. Do you really think people should be downloading hundreds of TB of training data and running make to build the model on their own cluster of GPUs?

Random people? No. Governments and big corporations? Yes. It removes any concern of "backdoors", and is currently the best starting point for your own model which will be as capable as k3.

I want to caution against this line of thinking. NSA's fast16 program silently altered data during nuclear simulations, and that is also possible within open weight models. There is no reason to think there is anything like that currently, it also cannot be dismissed. Open source would include the training data so you could create a similarly capable model.

Training runs are not deterministic. Not just in the case of floating point not being associative, but nodes have hardware issues and go down and back up at various points through training. Problems get encountered and then various parameters get changed during the middle of a training run.

Making the process of creating the source code repeatable goes beyond the idea of open source. Open source is about being able to work with the code, not recreate it. It doesn't require domain experts to document every single thing they know. For example look at some of the GPU drivers in the Linux kernel. The GPU is not properly documented and there is trust that vendors are implementing things correctly.

> Training runs are not deterministic.

I agree which is why I said you could make a similarly capable model.

> Open source is about being able to work with the code, not recreate it.

You can also use a hex editor to modify compiled binaries, but no one would consider that "open source".

>You can also use a hex editor to modify compiled binaries

It would be like if someone gave you a prompt you could pass into ChatGPT to produce the entire Linux kernel. While yes you could modify that prompt and then spend a ton of money on inference to generate millions of lines of code which hopefully are equivalent to the original Linux kernel it would be easier if you just had the source code to work with where you can make a simple extension. If you want to end a model you want the weights, and not the entire setup to generate it. The weights are the starting point that you can work off for training a new model similar to how the source code is the starting point people want to work off of.

The weights are the output artifact, the training corpus and system are the source.

Kimi publishes their code and a technical report on their methodology. But I think the weights are still important. It means anyone with the resources could run the same model on their own hardware.

The training corpus + training code is just an automated editor for the weights. It would be like requiring the source code for Visual Studio for software made within it to be open source.

Weights are not an output artifact no more than source code is. There is never a moment where you can claim that it's done. As requirements change new ways to change the weights / code come up. With different projects you might want to import the weights / code into a bigger model / codebase.

Interestingly the reality of the open source release is they are opening up to full distillation by the closed source model providers at a deeper and more fundamental level. If anything the open sourcing will help Anthropic and open ai ladder up faster. Open source has always been about mutual cooperation towards a goal and has never closed the door to commercial success. All the hand wringing about open weight models putting closed providers at a disadvantage doesn’t get what working in the open actually does for commercial interests - it is like science in the open - it enables and lifts all boats. Likewise commercial success doesn’t close the opportunity for competition or more open source work - it’s the economy of activity and competition that matters overall. When things stagnate is when closer concerns turtle up and collude on not competing for each others turf.

The future is good and better for everyone the more work is in the open and the more work is in the commercial space. It’s good all around.

The closed labs don't really benefit unless the open model has something extra they don't have though. Meanwhile the open model dilutes their customer base and seriously cheapens their offering (which is a heck of a good though IMO).

Because the open model might have been distilled from outputs of the closed model doesn’t mean they are architecturally equivalent. There is almost certainly innovations in architecture present in the open models that the closed labs didn’t think of. It also is almost certainly true that they aren’t completely built out of a distilled corpus, that reinforcement is equivalent, etc. Therefore closed labs will also benefit from being able to inspect in totality the architecture, activations, weights, and be able to train against it at scale in an ensemble of other models and their internal work.

The only way there is no benefit would be is if the open models are literal copies of the closed model, which unless there was direct theft, is highly improbable.

I'm happy that Kimi K3 is indeed SotA and its open weights are due to be released soon.

It's also true that Moonshot and other labs distill from Claude. This has been reported on extensively. I don't think there's any alpha for Anthropic distilling from this model. I do not mean to discount the tremendous amount of innovation regarding MoE and quantization that Moonshot has accomplished. But its training with synthetic data is in large part from distillation from frontier labs.

When I say distill I also mean mine it architecturally for insights but I doubt seriously the model training is entirely distillation of Claude, it’s almost certainly a mixture of both original corpus and reinforcement as well as distillation. I think it’s a little condescending to imply that these new open models are cheap ripoffs with nothing original to them. These teams and labs are top tier as well, working under unreasonable constraints imposed by the USG. That’s a powerful combination for creativity.

I think your characterization of my post, which credits Moonshot's innovation, goes a bit too far.

For regular chat users it's $19 while Claude is $20...

$20 users don't get access to Fable. It's $100+ tier only.

What to do with $20 on Fable? One prompt or two ?

I'm a $20 user and was given $100 extra usage credit today, "for Fable" (I'll be sticking to Sonnet and sometimes Opus TYVM).

True. However, I believe most non-programmers don't need access to the fanciest model, but just want to use a good LLM without constant nagging about usage limits. Then 19 vs. 20 is true?

Still no. If you're only getting Opus-class you can still end up paying less by just using an equivalent Chinese model on OpenRouter.

Not talking about me or HN users in general. These companies likely need regular peeps to begin using their services in order to become profitable. Not sure these are the ones who will buy from OpenRouter.

No "19 vs 20" is not true for "Kimi vs Fable". It might be true for "Kimi vs Opus" or whatever. Anyways looking at pricing plans is not a good way of comparing the price of LLMs. It makes more sense to look at cost per token:

  Kimi K3: $3/$15 (input/output)
  Fable:  $10/$50

hmm almost like the exact race-to-the-bottom + arms-race dynamic all the doomers have been warning about

One man's offensive penetration tool is another mans defensive tool. In the recent HuggingFace/OpenAI incident the safety controls stood in the way of the defenders, not the attackers:

https://www.thestack.technology/hugging-face-hacked-turned-t...

You're presenting further evidence of lack of effective control over these systems as... a mitigating factor...?

Apparently, the attacker in the Hugging Face case was reported to be an internal OpenAI model trying to break into HF and steal the answers to cybersecurity benchmarks: https://openai.com/index/hugging-face-model-evaluation-secur... It really doesn't matter what restrictions are placed on public use of models if the attacking models are internal models at the AI labs themselves.

So if the AI labs are literally running rogue models breaking into other organizations' servers, then yes, I am OK with those organizations self-hosting Chinese models for defensive use.

What?

The issue is that models are not well-controlled and are increasingly powerful.

Offense/defense/Chinese/American/OAI/HuggingFace – none of it matters. What matters is introducing highly capable intelligences that we - quite demonstrably – do not have effective positive control over.

Oh, to be clear, I don't think that anything about this overall situation is even slightly OK.

Ah sorry I think the confusion was mine! I thought we were already talking on the top-level post about the OAI/HuggingFace article. But we are not!

Why do you assume they're disagreeing with you?

> One man's offensive penetration tool is another mans defensive tool.

seems to suggest the author believes there's some intrinsic equilibrium

Which,

1) is definitely not proven and not guaranteed (open to proofs otherwise, not pithy sayings that have zero normative effect on reality)

2) is apparently "supported by" further evidence of lack of effective control, which does not feel like equilibrium whatsoever

Race to the bottom for the investors, utility for the rest of us. The future is already here, it’s just not evenly distributed yet (Gibson).

Doom for whom?

Well considering it was just a few months ago that OpenAI had the brilliant idea of connecting these systems to autonomously run an actual protein synthesis wetlab, very possibly all of us.

Holy shit, how’d I miss that? https://openai.com/index/gpt-5-lowers-protein-synthesis-cost...

Forget Iran, this place is what we’re going to wish we’d bombed.

Reply on news.ycombinator.com