Classification models have been around for literally almost a century at this point. I think it's safe to say they are a proven technology.
The only thing that makes Jev and the likes particularly interesting is that it is a general purpose classifier. In the past, classification tasks meant training a new model to solve your problem. Now you can just use an off the shelf general purpose model and hit the ground running.
It is in that sense not integrated with the coding agent. It's just that some things cannot be done with bash alone, at least not as trivially. So if you were asking Pi to utilize Jev, it would not really have the right tools available to make sense of it, even though pi-ai, the underlying library, can make requests to it.
Codemode as a mechanism can expose non LLM functionality to the coding agent. In that sense, Pi does not have a tool for Jev or other classifiers. It just now makes it easier for the agent to utilize it in the same way as it's otherwise quite creative in using bash.
>it would not really have the right tools available
The point of Pi is that the user can tell the agent to improve itself and give it the tools it does need. The minimalism comes from the user creating what they need instead of the maintainers trying to support everything for the users. The fact that it doesn't have everything the user needs out of the box is intentional.
> The point of Pi is that the user can tell the agent to improve itself and give it the tools it does need.
The point of Pi is to be minimal but also follow what the models need. We were pretty outspoken that models need code execution, and that's why Pi to this day has a very small set of tools available. However as more and more training with these models abstracts even over toolcalls themselves with code mode and similar things, it requires changes to Pi.
With Pi the agent edits agent itself. That's one of the reasons it's written in typescript, to make such iteration fast. Going even lower, into the language runtime or operating system shouldn't be necessary but technically also possible.
https://america.gov isn't even calling artificial intelligence "super intelligence". Trump just says words. People should stop trying to find meaning in them.
> It is therefore the policy of my Administration that, to the maximum extent permitted by law, the executive branch shall use the terms “Super Intelligence” and “SI” in place of “Artificial Intelligence” and “AI” and will not acknowledge the usage of “Artificial Intelligence” and “AI” in any applicable setting.
My colleagues at NIST just got a memo directing them to refer to AI as “SI” so it does seem to be filtering down to the agencies. This is a problem because “SI” is also a classification marking.
> This is a problem because “SI” is also a classification marking.
They can just make up another name to redefine this classification which people are already used to. How about SIBISI? This follows the transformation they instituted a'la kilobyte -> kibibyte.
I know you said /s but I'm feeling all serious or something this morning, so ...
- The US government's official policy is metric (and has been for something like 50 years now).
- The population uses a mix of metric and US customary units, not imperial.
UK is kinda similar, I suppose, but their typical usage of metric/imperial tilts more heavily in favor of metric. I wonder how many other countries are also still doing mixed measurement systems at this point?
I get the sarcastic intent in your comment; but it's more useful to focus on what Trump does than the things he says.
One of his strategies is to say unbelievable things like he's gonna revoke Rosie O'Donnell's citizenship or something weird like "superior intelligence", and the news will cover that instead of whichever agencies he's getting or how much he appears in the Epstein files.
"People" don't have that luxury. Especially internationally.
From an international perspective, America is Trump, in a way that is far more acute than even when it was Obama or GWB. Back then you could understand POTUS as sitting on top of an enormous apparatus that didn't just act out his whims; in the current situation we have to treat him as a sort of dimwit emperor.
I think you can say the same about how it is viewed domestically, as well.
Perhaps the right wing always thought that the government was the personal weapon of the sitting president, and that is why they are so comfortable with actually making it so, but the left wing has predominantly seen it as an enormous bureaucracy governed by policy and laws that only changed relatively slowly, and not at the whim of an individual.
Now when I see data and the source is the federal government, my reaction is "this is probably fake, or at least biased heavily." Makes me a little sad to have that reaction, e.g. I used to like going to the NIH web site to find information, but now I assume it has been curated to meet an ideological narrative.
I also do not have a lot of confidence that this situation will be a one-off aberration and the democrats will bring sanity back and implement better safeguards against the next demagogue. On the contrary, I think it is very plausible they will see this as an opportunity they had not dreamed of before.
It's annoying, but that's been pretty common for the last decade. It's just like how everyone uses "REST" to mean "JSON RPC". In the end, we all basically know what the other ones are talking about, so it's pretty pointless to get bogged down in semantics.
It's common when talking about web services, where the kind of API is generally unambiguous. The problem here is we're discussing different kinds of APIs (shell commands, MCP, REST / JSON RPC) and so calling one of those kinds of APIs "API" is very confusing.
I'm not trying to be a dick, but it's pretty hilarious that we made these two posts at the exact same time [0]. I totally get what you're saying, but I don't understand the motivation behind it. A properly designed HTTP API is what Roy Fielding was discussing in his dissertation, but he was obviously talking about HTML pages full of hyperlinks. At the end of the day, does it really matter much if the definition is only 80% correct if the majority of the population understands the basic gist of the conversation?
If you can build the ultimate evolvable client, then you can collapse all UI into a single client.
Basically Roy Fielding told us to build an API that can only be operated by a human like intelligence, nobody could implement that because such an intelligence did not exist (hence the switch to imperfect HTTP APIs), and now that LLMs are a thing, said human like intelligence exists. This means Roy Fielding wasn't wrong, he was 25 years too early and ironically people should be building real REST APIs in the pendantic academic sense from today on and not the "pragmatic" HTTP API.
MCP is not strictly an API; it's a protocol shape (thus the "P" for Protocol) for delivering an API.
The API surface exists in the MCP payload.
Whereas application teams previously would have focused on APIs for external access, they now have to focus on MCP entry points (often to those same APIs, but with a different shape).
> I've yet to successfully convince one to tell me when it knows something.
This is a near daily experience for me when using a coding harness. I will ask it for some favts about the environment, and it will continuously explore the environment until it exhausts reasonable exploration, or it finds the facts.
It's long, but I recommend watching this video[0]. I believe the person being interviewed has one of the most lucid understanding of the potential futures regarding if AI is allowed to continue or not.
This show is trash are you going to recommend I get medical advice from Joe Rogan next?
Non domain expert sitting there making Pikachu face while deeply questionable alleged domain expert rambles on and on for hours is such a low form of information.
Like please stop watching this. It's the same trash format the guy just talks in a soft English accent so his mediocrity is less obvious. You can do better.
What happened to debates? What's up with soft ball interviews and people going on and on for hours unchallenged? What's the point of listening to someone not willing to subject themselves to cross examination?
> I haven't seen one convincing model of AI existential risk. Can anyone offer one?
I apologize for answering. This is the only episode of this podcast I have ever watched, and I have watched it four times now. The person being interviewed knows what they're talking about. You would learn from them.
As much as I hate appeal-to-authority based arguments, it's basically the only way I can see how to answer this without just totally rehashing the things he says in the video. He was an AI forecaster at OpenAI for several years. (I think 2019? to 2024). He quit on the grounds that OpenAI (as well as the rest of the industry) were / are throttling towards RSI irresponsibly, and even refused to sign an anti disparagement clause on exit at the risk of losing 2 million dollars (80% of his family's net worth) so he could talk freely about OpenAI. Now he runs a non-profit that does AI forecasting and advocacy.
Sometimes the most interesting people are people with ideas outside of the same groupthink circle. Everyone who has changed society has been called a quack by a square.
Dr. Anthony Fauci invoked his Fifth Amendment right against self-incrimination more than 100 times and declined to answer questions during a heated July 29, 2026, Senate Homeland Security and Governmental Affairs Committee hearing.
He was wrong about so many things and now refuses to talk. Also got the first pre pardon for crimes undiscovered against him.
Debates can be better than CNN gotcha questions, you know that right?
At the end of the day a debate is only as good as the interlocutors. I wouldn't advocate spectating two people screaming at each other either.
It's just sad that I can't even point to an instructive example here. This is part of my frustration. I don't think I've seen a good contemporary debate in a long time.
I think the reason covid became politicized is the same individualistic vs communitarian tension in all other american politics.
Individualistic: vaccines don't help the young
Communitarian: vaccinating everybody creates herd immunity that protects the community
Individualistic: Masks don't protect the wearer
Communitarian: Masks reduce the virus the wearer breathes out reducing the harm to others
On ivermectin Rogan was just wrong.
On lab-leak, Rogan pushed it hard based only on circumstantial evidence while Fauci thought natural spillover was more likely but didn't rule out lab leak. I don't think we'll ever know for certain.
Who should I trust, the extremely political Trump white house that never misses a chance to create division and attack perceived enemies. Or the scientists and even US intelligence agencies that disagree with each other and continue to debate the topic?
From my understanding this intel came from US intelligence agencies. I don't think scientists are against this viewpoint. In fact many spoke out against the pengina / wwt market theory.
Not really related to the central point, by but I couldn't help but get caught up by
> Generally, models intended to be run locally will be much smaller, such as Muse Glimmer or Qwen3 Coder.
That is such an interesting set of models to use as examples here. One being essentially obsolete on release a month ago, and the other being completely ancient in LLM time. I really wonder how they landed on those two.
Don't cut the quote mid-sentence. Here's the full quote:
> note: this is a parody blog post, see these links for better/more complete open implementations of Jev: OpenJev, openjev-sglang, and OpenJev on DiffusionGemma.
Are OpenJev, openjev-sglang, and OpenJev on DiffusionGemma using Qwen3-0.6B-Q8_0.gguf, or did you just want to emphasize the part that was unrelated to your previous response?
> To me it clearly means "releasing frontier models at any pace less than as fast as possible".
It also seems to misimply that the "frontier" that they release is the same as the frontier behind closed doors. Who's to say they are not throttling full speed towards RSI privately while pacing their public releases?
Not the same person but... nothing. Haiku just hasn't been an interesting model for a long time. If you want cheap and fast, there are lots of options that are simultaneously cheaper, faster, and capable than Haiku.
The only thing that makes Jev and the likes particularly interesting is that it is a general purpose classifier. In the past, classification tasks meant training a new model to solve your problem. Now you can just use an off the shelf general purpose model and hit the ground running.
reply