In my anecdotal experience, it is not. Same model, opus, works better in 3P harnesses such as Factory Droid or Amp.
Claude code, on the other hand, is the most subsidized one, both for consumers (through max subscription) and for enterprises (token discounts). It is also heavily optimized for cost, specially token caching and reduced thinking, at the expense of quality.
Previously cheapest Mac-mini was $599, M4 with 16GB RAM and 256 SSD.
Apple removed the 256GB Option, so now the cheapest Mac-mini is now has 512GB SSD and MSRP of $799.
The Openclaw and personalized/proactive agents caused a sudden surge on Mac-Minis.
It remains unbelievable to me that people were rushing to buy an entire Mac Mini to run obscenely bloated software that ought to have been able to run on a Raspberry Pi.
i just want to know where emdash came from, as it is quite rare to see it on the public internet, so it must have been synthetically added to the dataset.
Emdash is very common in academic journals and professional writing. I remember my English professor in the early 2000s encouraging us to use it, it has a unique role in interrupting a sentence. Thoughtfully used, it conveys a little more editorial effort, since there is no dedicated key on the keyboard. It was disappointing to see it become associated with AI output.
The very simplified answer is that the models are first trained on everything and then are later trained more heavily on golden samples with perfect grammar, spelling, etc..
Other than things other comments already mention, let's not forget that Microsoft Word auto-corrects "--" to em-dash, and so does (apparently - haven't checked myself) Outlook, Apple Pages, Notes and Mail. There's probably bunch of other such software (I vaguely recall Wordpress doing annoying auto-typography on me, some 15 years ago or so).
I think it's because of Wordpress sites, as their titles often have them and the editor automatically turns things into them. A large part of the Internet has been powered by WP.
It has been rare. It's common now, even in meaningful human texts. (I know because I detest the correct usage without spaces, t looks wrong.) One of the ways AI is shaping our minds.
May i ask what was the decision process behind this? What was the benefit of open-sourcing warp, as it is already a mature and established product. Also did devin cli had any impact on the decision to open-source warp?
Also how does a repo gets 29k starts in matter of 2 hours?
But the tl;dr is that I actually think we can build a better product, more quickly if we build it with our community + agents. I also think it's a unique product that I hope developers get a bunch of value from being able to customize and help improve. Our business is now mostly around agents and orchestration through Oz (https://oz.dev), so opening up the client and terminal felt natural.
The big thing for the "why now" though was the agent management piece.
Wrt the github stars, we had an issues-only repo prior and already had a significant number of stars before OSS today.
> Barring them from leaving the country feels a bit sinister for people who haven't been accused of committing any crimes.
Pure speculation on my part, but i would be surprised if China didn't have our equivalent of export control laws, not difficult to fabricate a crime and pin it on founders.
Yr parent is new to standard China legal mechanisms and you pivoted off of that to invent a chain of stuff that isn’t real. Are they unfamiliar to us? Yes. But it’s worth speaking to whether the speculation is rational.
They do have export control laws and such, but based on current and past behavior China’s Communist Party doesn’t need those laws to disappear people or create crimes and then make people guilty of said crimes.
Worth mentioning though that this is not how America functions, nor our rule of law.
Not much, none of those cases from the US resulted in disappearing the founders. The US is a nation of laws, no matter how imperfect. Stark contrast to the CCP.
At the end of the day, the process itself, years of investigation, millions in legal fees, frozen assets, destroyed careers is often the punishment regardless of whether charges stick or convictions hold up.
The US is a democracy, and people are given many procedural and substantive rights, even Guantanamo detainees (we can argue if Boumediene had any practical effect, but we wouldn't have seen the same from China).
But Americans are under the impression that what the world sees is what they mostly see -- the domestic side. And to a certain extent, they do thanks to its cultural influence. This democracy/rule of law, however, is completely absent in way it behaves outside its borders and it's now clearer than ever to everyone that the US is the biggest source of instability in the world. More than Russia. Certainly more than China.
Then you probably are not fit to comment on this matter.
I'm sorry to be that blunt but if you don't understand the value of rule of law, the difference in incentives, the consequences of separation of powers, I can't even grasp what kind of perspective you can build. It's genuinely baffling to me.
Quite the opposite, AI should work longer and interrupt the human less often (Tokens are cheap, interruptions are expensive). So we want to push the agent's horizon to infinite, now when they do make interruptions (e.g. Creating diff) it will be larger chunks and more complex, so these summaries are actually quite useful.
You can and should have both smaller chunks and a larger time horizon. The AI should output code in a format that's easy to review.
The Linux kernel submission guidelines are one thing you can feed the AI to guide that. Work is submitted as a patch set, and each patch in the set must be small and self contained. Many patch sets are over 50 patches.
Claude code, on the other hand, is the most subsidized one, both for consumers (through max subscription) and for enterprises (token discounts). It is also heavily optimized for cost, specially token caching and reduced thinking, at the expense of quality.
reply