> The EUIPO found that the word "open" would be understood by the relevant public as meaning freely accessible, while the combination with "AI" (artificial intelligence) would be interpreted as referring to products based on openly accessible artificial intelligence.
> for certain software and information technology goods and services, the term is purely descriptive and therefore lacks the distinctiveness required for trademark protection
> Opus 4.8 built in Claude Code; GLM-5.2 built in Pi over OpenRouter.
It would be more interesting and accurate to see the comparison on the same harness if the intent is to compare the frontier models.
Pi is relatively new and does not have many features built-in compared to Claude Code. It was chosen intentionally this way as Pi's goal is not to create a bloat builtin of tools most don't use but to allow the users to customize to fit their need -- similar to Neovim vs IDE.
The end-user "vibe coding" experience is *heavily* swayed by the harness because prompt effectively drives how a model outputs an answer.
I heavily using Deepseek V4 Pro for a personal project because I cannot afford Opus, and spent ~1B token last two weeks for just $40 which would've costed ~$1300 using Opus 4.8. Realistically Opus cost will be lower assuming more "intelligent" model would've produced less code with fewer conversation but I doubt it'll be cheaper than ~$500.
I'm curious to know how they can they offer at such a cheap price. Some say it's electricity surplus in China and/or government subsidy. It'll be a very interesting read if there's an extensive study on their economics.
I want to emphasize "unknowingly" part. Once data is uploaded to an entity, there's no guarantee they'll manage is properly.
User agreements change constantly, engineers make mistake, firms get liquidated and data might get sold, and most importantly, as a former employee of social media firms, what the firms say about the user privacy publicly is very different inside.
If Anthropic really found a model that's so powerful no one has, why don't they use it for themselves to create things no one can and accumulate unimaginable wealth?
The problem with photo since the birth of social media is that it's permanently stored in the internet, literally.
Photos used to be personal and (mostly) temporary. I may take a photo in public, develop, then share with the close ones and store in the photo book. Photo may be somehow passed onto others but likely thrown away eventually when I become less of importance to them, and it'll worn out.
With photos now uploaded to social media or the "cloud", they exist permanently as a means of backups, sold to 3rd party (knowingly or unknowingly) analyzed to "improve the experience of the platform".
Got curious and ran a similar prompt with DeepSeek v4 Pro w/ OpenCode
No idea what's going on here but agent tested a bunch of stuff. Then I asked to build a wheel so I can run the command you noted above and it appears to pass
I don't quite understand the intent of such article other than to promote themselves given an odd timing that the company is planning on going public, so I can only conclude that this is just part of the IPO roadshow.
LLMs certainly have made significant changes to our lives, but I haven't yet to see any extraordinary improvement it brought to me which makes me skeptical about their claims.
_if_ it solves many of our problems of great magnitude, why haven't Anthropic used it to solve significant problems we, humans, face? Cancer, Alzheimer's, education, finding new materials, fission power plant, etc.
I believe surveillance can only work in society at a very small scale.
Surveillance == Information == Power.
Who then, manages the surveillance? You? Me? A government?
How do I trust them?
*Trust* is the fundamental problem with surveillance.
An innocent person was killed on a street. Surveillance camera on every corner of the street would've easily caught the murderer.
Who manages the recording? An individual? Entity? A government? How do I know they aren't face-tagging and selling the data?
It's always about trust.
An interesting case to consider is Seoul, South Korea which actually has cameras almost on very corner of the street. In my view, general population prefers streets recorded to avoid aforementioned situation, and it can only work if people have trust in the government. In the US contrary, my view is that the disconnect between the government and the people are too wide to have such trust.