leobuskin

There’s no way the current safeguards affected by “trust” score in dynamic way (and the score is fluctuating), more like the tiers of trust are being proxied through different policies - it’s a separate story. But I observe the same “cyber” issues working with Linux kernel and drivers as people who went through CVP/T1 (which is wild, to be honest). Hypervisor interactions trigger immediate “cyber” response, it’s impossible to work with Opus 5.5 doing anything low-level with custom or third-party devices. Only 4.8 can assist, and it’s a huge difference in code quality and approaches.

Anthropic had/has a great opportunity to equalize the landscape, but they are choosing the Cyberpunk 2077 branch for some reason.

US/EU here (both, due to traveling), same story with anything related to modifying existing weights or SotA approaches in training/research, it's not a geolocation issue. The safeguards don't check your IP (but pretty soon will check someone's IP [pun intended])

It is required. The first week of release I told here how GLM 5.3 was happy to redteam, it’s not anymore. Models modification, offensive cyber, exploits - requires significantly bigger prompt/scenario engineering, and still hits refusals.

And those services are oversold and underdone. 30 tk/s on Refuseless, huge TTFT, looks like a damaged quant (it’s NVFP4 and you see wild random stuff periodically, including foreign languages). Most of them using third-party quants (dealignai?)

If someone here wants to rent 8x B300 together to avoid problems all those services have - ping me via email. I don’t want to sell, I want to work.

X is under Cloudflare (and according to some of my probes may even use some parts of CF stack)

do you know what kind of mixers people use nowadays for such transfers?

Why would you prefer WSL over native Linux? What's the rationale here?

That's totally fine, I'm pretty sure people continue to use JB by inertia, requirements, personal preferences, lack of alternatives, inability to introduce some, or whatever is the reason behind. It's all for good. Most (I'd even say all) people I know jumped off paid JB products within the last two years (C#, Rust, Python, Web stacks), even being evangelists previously (I was religiously protecting and advertising JB in all possible holywars). And in 2023 we already were betting if there's a place for JB "as-is" in the current/future landscape. Too slow, too enterprisish.

CC + my own custom tools, I don't see any reason to pay for devtools in 2026 (same as trying to monetize them, or even just share/distribute). My first custom solution was built on top of JB's core in early 2024, btw, but not any more. JetBrains "suite" doesn't really lock you in, it's more like "aura" of the product itself that made me attached for almost 13 years (I never was able to switch, I still think VS Code is so ugly and counter-intuitive in comparison with JB).

It's purely management fault, it happened around 2022-2023. They had all the chances to become a great agentic IDE/harness at some point, if they would start in time, support plugin developers, allocate a proper team instead of 2 people and some semi-random internships. A lot of users were questioning this publicly and privately with them, we all knew how it will end this way. JetBrains lost, no one wants their Junie in 2026 (and didn't want in 2024-2025, whatever was the name of the integration at that time, it was awful). I was on Ultimate Plan since 2012, appx. I quit entirely in 2024. They asked to answer a few surveys over the last few years - even the questions there were far from reality. RIP, JetBrains, it's mostly unrecoverable at this point.

It’s two different dimensions in this case. VCS doesn’t preserve your local history (in JetBrains’ terms) of changes (and it wasn’t meant to by design)

Astra had an explicit medium/xhigh levels, Fable - just Fable. What reasoning level was used? Why not multiple were tested?

I’ve scrolled the article, but haven’t noticed any remarks about Fable’s levels.

An example with manual combinatorial exclusion in “not_for” field made me cry, this is a wild hybrid of code logic, textual definitions, and AI blackbox. It’s a cool idea, but the “glue” layer is too boilerplate-ish

mmm, someone decided to reuse `zookeeper` name, seriously?

Heretic?

That’s the first model that fits CC as it’s own, zero issues, but probably ZCode or whatever z.ai’s cli is.

On $80 plan 200M tokens ~20% of weekly usage

It’s comparable to Anthropic usage, to be honest. 2x GLM agents ate 18% of weekly usage on this mid-tier plan within ~8 hrs (non-stop work, a lot of tool calls, appx 4 compactions each), I think. I didn’t make a proper statistics snapshot, sorry.

In what scenarios?

I’ve tried a bunch of them, and I seriously do not understand these recommendations. It was a rough road and a steep hill, but right now CC is absolutely the best harness on the market, as for me, whatever top tier model is under the hood (mostly, some of them, like DeepSeek, don’t fit CC at all).

Fable wasn’t trained on github commits, let’s be fair.

Impossible with source code, possible to bypass with app/site

Exactly my CoT! I hope z.ai won’t change this behavior after training it on our input the same way as Anthropic did (shame on you, folks, seriously)

I bought $18 GLM official subscription yesterday (5.2, but new model version was already leaking on some docs), set it up with Claude Code harness... and I’ve bumped to $80 plan almost immediately. It’s the first model that agreed on a proper security research (red team scenario), executed it seamlessly, including 0-days in WP plugins, RCE, 6.8 kernel exploit adaptation, etc - while playing against another GLM agent as a defender (following HF story)!

I understand that such models can be used by malicious actors, but it’s fair to have it publicly available (and play on your side in case of emergency). This is what changes the world in a better way, I think, not the guardrails.

Nope, they want shared artifact(s) between web and desktop (not sure what do they have in mobile apps) clients (Chat/Cowork/Code/CC-cli), so TS is the way to go. It’s a reasonable choice made by engineers.

Is it some Claude Team/Enterprise only problematic? I'm using two 20x Max accounts almost non-stop (Fable/Opus) for 1.5 years at this point, zero issues with both client and infra sides (from US and in travels). When I'm reading such messages it feels like either I'm lucky or it's a part of some campaign.

If I understand correctly, time spent within this interstellar ship reference frame would be significantly less than 75k years. And at 1c speed, 48 years on Earth would be ~2 years of "flight" for the passengers, correct me if I'm wrong.

Is there a way to turn options-based AskUserQuestion off? I couldn't find it at all, and the "options" selector is the most annoying thing in CC for me, plain-text is the only way, everything else is distracting. I know, I can use `n` (sometimes) or cancel it, but both bring more pain that just regular communication (cancellation does one more chat step + requires an extra action, same with `n`)

That's not a bailout for AI labs, I meant the "bailout" for Salesforce and others. There's absolutely no place for them in the world where we have Fable+ models. For many of them. Most of them (we just didn't get this feeling yet). Someone just trying to maintain the old world order, that's all. I don't think US economy would fail if those absolutely useless giants would go down.

@anthropic, can you finally add $800-$1000 per month plan and allow us to work instead of tracking your weekly changes and dramas? I think, we (individuals, small-medium biz, first of all) did our best to help you train the model like Fable. Enterprise-level lockdown (and API costs define this) is... unfair? I mean, we all knew that you all will just use us, but it's AI, right? For people, right? Right?

The only reason this is happening -> someone (US gov?) decided that it's time to bail out those who would inevitably die within a year or two otherwise, middlemen.