boshalfoshal

1. I don't think it'll get acquired by a frontier lab. Frontier labs already focus on robotics with enough effort that acquiring pure-software research IP from a company like this is not very useful. OpenAI has way more compute and researchers than these guys, and there likely isn't very much dream is doing on the model side thats different from OpenAI, or what they are capable of doing.

2. However to argue against basically any other robotics startup, I think they would only be acquired if it was a robotics SW company _with real customer distribution_/revenue, or a robotics company focusing on HW that is worth acquiring the IP for. The problem with the latter is that it is also a huge cost, and that cost is only worth it if the ML side actually catches up to the HW. I don't think a frontier lab would want to sustain a full physical supply chain for a non-revenue additive product. And frankly, most robotics companies have no meaningful distribution/revenue or SW/HW IP that is worth acquiring, considering the cost.

I disagree with the claim that its sticky, but this is the wrong response. Reductive and doesn't take into account the obvious things (UX, network effects, barrier to entry to use the product). All 3 of these are much better with Muse than they with are similar products.

The notion of human-readable/human-safe programming languages will be irrelevant in a few years, not just C++. They may use Rust, Zig, Lean, or maybe some other LLM-generated hyper-optimized intermediate language with a strong type system and compiler, but chances are, we will read less and less of it, until we basically never read code at all.

I know this is somewhat of a meme now but I do think the predominant language will be "typing/speaking to a coding agent in your native language."

I mean Rust definitely has a better tradeoff than Carbon in this case, re readability/verifiability by a person (and sufficiently good internet training data).

I personally think that you _could_ use an LLM to catch these types of boundary case errors without having to port the _entire_ C++ codebase to Rust, but maybe pre-emptively porting to Rust now can catch some of these cases for cheaper than doing a full LLM sweep. Also more cynically, its a good benchmark lol.

I guess if you really believe in curve of LLM capabilities you should just use a language that has the best performance, safety, flexibility, and extensibility, since in the limit few/no people will actually read the code anyway. I think this ends up being Rust.

There is 0 practicality in inventing an entirely new coding language that only one company uses, and you have to teach it to thousands of new engineers. Rust exists and fits the job totally fine and is used in more places and has actual support outside of a single entity (i.e you can actually hire people that feasibly know the language).

It was clearly done because some PL guys at google really wanted to make a new cool language and Google was the perfect place to incubate it without it getting axed. Probably got a couple of promos out of it too. This is clearly not the best use of time or money, but I guess if you're google you have so much of both it probably doesn't really make a dent, and you can keep a few very smart people happy with shiny new projects.

Also, LLMs being used for a large portion of coding nowadays sort of remove the need for these types of languages, IMO. They make less "silly" bugs (both logical and structural) that languages like this are meant to catch, and they are much better at languages that are better represented in the training corpus. This somewhat obviates the need for very niche "type/dummy-safe" languages like carbon (and even rust/zig, imo). So even if you did want to use Carbon, you'd likely have to bootstrap a decent amount of your own "good" carbon code to post train an LLM, and even then, it likely won't have that big of a gain vs just having an LLM write C++ or even Rust. If you are a company that still reviews code, you should just have an LLM code in a language most people can understand anyway to make verifiability tractable.

Right, and lets ignore the fact that Meta rocketed up 50% in a month, and the rest of the market is also down ~1% because of crude going up.

-4% means nothing in context.

And I'm pretty sure this "29-30 year old who was merely acquihired" probably knows more than these veteran/boomer execs about how to execute a consumer AI product strategy.

"Chief AI officer" does not mean he leads AI research efforts, its just a title. He just takes care of hiring and product direction, both of which I'd say hes done well in. They obviously have actual researchers working on models.

I'd go as far as to say Alexandr Wang can do this better than most other leaders. Young people should be in charge of products, they generally have a better sense of what resonates with consumers, and Meta is a consumer company at the end of the day.

We went from open source Chinese models eating _all_ of frontier labs lunch, to them eating some small extremely niche market of frontier labs.

I don't think this will happen, just as Chinese models did not eat Anthropic/OpenAI's lunch on top tier intelligence. The market is way too niche, and the parties already interested in the capabilities behind safeguards are likely already partnered with Anthropic to get around those with Mythos-class models (see project glasswing for cybersec).

Ok and they just released Opus 5.5 today which is at or above Fable 5.1 level capability in many tasks, but costs a fraction of the cost and is faster? And this is what, like 3 months after Fable was first released, and 3 weeks after Fable 5.1?

Please tell me models are not improving at breakneck speed.

Clearly no one here has heard of Occams Razor before :P

Nerd sniping the hackernews/twitter crowd with "large scale transformer-based language model alternatives."

People don't want to believe something as unsatisfying as "Scaling up LLMs" can yield something as profound as AGI/be useful, and just hope that literally anything else can take their mindshare away, and this just happens to be the new rage. Along with clearly-not-frontier-level open source models, non-transformer based architectures, etc.