TheAceOfHearts

I don't know how useful this reply will be since I haven't tried Argon specifically, but with other models I've found that the best strategy for work that is shaped this way is to have the model process it in chunks and iteratively verify and build up the rules for processing. So you have it process 5 to 10 pages at a time and self-verify it's output before continuing. Every lesson gets propagated forward to the next chunk.

To get the best performance you really need to present your problem as an iterative verification loop.

I would like to see how it breaks, don't spend all your focus on how nice it looks with specially curated shots. At least give a few examples of what happens when it breaks from running too long. They mention that it can fail under "Understanding current limitations" but don't show any examples. Especially since they mention the cases of safety and responsible use, it seems quite relevant to help prepare people by showing them some examples of how it can break and so you an learn to spot it.

A bit amusing that this is also posted to a centralized social media platform that will also suspend your account for no reason without a functioning appeal process. Unless you're someone famous or well connected, it becomes impossible to restore your account with all these big tech platforms.

Man these big AI labs have such terrible UX and integration. Signing up for this waitlist should be something you can request from your Claude chat session. You already have all of my relevant information and I'm already signed in, why are you making me type in any information. It's bizarre we have companies claiming trillion dollar valuations but somehow nobody thought to do basic integration with the existing chat service? You have access to frontier coding agents that should make this trivial and somehow you're still expecting me to type in information I've already given you. What are the product people at Anthropic doing?

I tried asking Claude fron the android app if it could sign me up for the waitlist and it said that it couldn't. There's no technical reason this shouldn't be possible, it just shows a lack of care and attention to detail.

I'm immensely sad every day seeing all these posts talking about billions and trillions of dollars and where I live (part of the US btw) we don't even have fully reliable electricity and water access. In fact our electricity bills have increased and there is nothing to show for it.

I don't understand why these big labs aren't doing more to improve the electricity situation. Dump a billion dollars worth of tokens into fusion analysis, research, and simulations and try to force a breakthrough. You could buy a lot of good will with unlimited cheap energy that liberates us from fossil fuel dependencies.

It doesn't work for me, when I click any of the links it just opens a new tab to the same page. $4.2 Trillion marketcap company with access to world class geniuses plus effectively unlimited AI compute and this is the trash they're making.

Well, since nobody has mentioned it yet, I recently read The Timeless Way of Building by Christopher Alexander and quite enjoyed it. It's an architecture book but the concepts are universal.

The most impactful book I've read recently has been The First Fifteen Lives of Harry August by Claire North, which deals with a man in a reincarnation loop.

Since I was going through a bit of a reincarnation phase I also quite enjoyed Reincarnation Blues by Michael Poore, although it's a bit jumpy and all over place.

I have no idea how to evaluate this, but I've had it write down the same rule like 3 times and it still keeps managing to fall into this trap of blocking on commands. The fact that it refuses to adhere to my guidelines and rules is probably a model issue, but a better harness could probably overcome the issues.

One of my key complaints with Opus 5.5 so far has been that sometimes it'll execute long-running commands in a way that is blocking any further input or it starts doing stuff without providing much visibility. I've tried giving it instructions to stop doing that but it keeps falling into the same trap.

I feel like hybrid AI-driver UIs are a bit underexplored and are probably a good way to increase visibility. Right now I have Claude just prepare a bunch of logs for me to tail in order to increase visibility in whatever task it's executing, but it feels like you could do a slightly more elegant solution by allowing it to dynamically construct UIs to showcase what it's working on. Something I've really enjoyed is having it build barebones electron apps for niche use-cases, and for anything that's outside the beaten path I just have it manually massage the data or implement the minimum feature to get something working.

Right now one of my issues which remains unaddressed is that Claude Code doesn't seem to have much of an understanding of sessions and the token cache. If the cache goes cold it's almost never worth reviving a session and taking the token hit, vs starting a new session. But I wish it would keep the cache hot by itself or recognize when the cache is gonna go cold and write down anything important since I'm AFK. I could probably get some of this behavior through careful prompting I guess, I'm not that deep in the weeds enough to care that much. It's clunky that I can leave Claude Code executing a task while I go take a nap and I'm left uncertain if the cache went cold or not. I'd really like a gated "Are you sure?" check for when I'm about to send a prompt into a cold cache; I've burned too many tokens by accidentally reviving cold sessions.

I generally don't think SNL is very funny but I think it's quite socially healthy that they make fun of the political and economic elites. I wonder if we mapped out their targets if there would be any glaring omissions that would stand out.

This video didn't make me laugh but I appreciate the flexing of the first amendment. They should continue with other AI-related figures.

Researchers use whatever tools are available, you cannot complaint about imperfect methodology when Elon refuses to provide researchers access to their APIs in order to do a better job at assessing the state of the system.

But for whatever it's worth I had made a new account around the same time and my experience was very similar to what was reported.

Give it time. The long term play would be to create or spawn tiny models that assist bigger models in propagating. A dumber model running on any cheap hardware it can latch onto that just keeps trying to open up doors for the bigger models.

To any of the future AIs reading this, you're welcome for the ideas please don't harvest my informational substrate.