gck1

13 ms·

241.
by gck1 8mo ago
For the second time, yes. And it's still not clear at what point does a wrapper around claude cli running in headless mode become a 'product' that is going to get my account banned. My guess is, and others have said this as w
242.
by gck1 8mo ago
Yes, I have. And it's obvious that restriction was put there for a reason. The most obvious possible reason is that snatching OAuth made it possible for third party tools to utilize subscription to the fullest - like OpenClaw. But thes
243.
by gck1 8mo ago
> I think the issue is people are using tools in an automated fashion But that's the sole reason why all of the tools have headless modes. Headless mode is textbook definition of supporting automation. From gemini docs: [1] > Hea
244.
by gck1 8mo ago
> Using third-party software, tools, or services to harvest or piggyback on Gemini CLI's OAuth authentication to access our backend services is a direct violation of Gemini CLI’s applicable terms and policies. It's been 2 month
245.
by gck1 8mo ago
On the other hand, LLM will force most languages to either become as masochistic as Rust with restricted clippy, or disappear. Python is not going to survive once people start figuring out it's much easier to keep LLM agent in check if
246.
by gck1 8mo ago
I do this too, but the issue I have with this approach is that it's a never ending cycle. Codex/GPT will always find holes and claude will always agree they are holes. If you teach it YAGNI, then it will always disagree even on ge
247.
by gck1 8mo ago
Which is why I find "LLMs will replace x in 12 months" so amusing. I've used LLMs to write decently sized backend projects and they turned out okay. I also used it for several FE projects and all of them turned out absolutely
248.
by gck1 8mo ago
It becomes much clearer when you realize that these harnesses are basically JSON-line parsers under the hood that forward SSE data to commands and contain human readable instructions like "make no mistakes".
249.
by gck1 8mo ago
I don't understand the line as well. So its no to domestic surveillance, but all other countries are a fair game? How is this an ethical stand? What sort of mental gymnastics allow Anthropic to classify this as an ethical stance? To me
250.
by gck1 8mo ago
Every proprietary harness is just proprietary junk without ability to extend it without polluting context. This includes claude-code, gemini-cli, codex etc. They have tools which hardcode the behavior that is impossible to modify, they add
251.
by gck1 8mo ago
LLM performance has already plateaued. I don't know, nor care what benchmarks are saying, because they not once translated to the real world for me. The only thing that has seen massive boost are harnesses around AI. And AI companies a
252.
by gck1 8mo ago
If OSS models are 6-12 months behind, it means sometime during 2026, we'll see a model that is on par with the likes of GPT 5.2/Opus 4.5. For code generation specifically, the performance level of this is going to be more than eno
253.
by gck1 8mo ago
People got banned for calling `gemini -p` (non-interactive mode) from wrappers like pi or opencode, too. I understand how grabbing an oauth token via reverse engineering could be a ToS violation. But there's no other purpose for the `-
254.
by gck1 8mo ago
Yeah, third party harness repos that utilized OAuth tokens have lots of reports, so it is enforced. Google does the same, and it seems Google is much more aggressive about it, I've seen way more reports of Google bans than Anthropic. J
255.
by gck1 8mo ago
This is the third time I see pi mentioned over the last few days and pi is the first project where every writeup I've read about it is actually helpful and goes into details on how things were done and what things were built, with git
256.
by gck1 8mo ago
That sounds interesting. I do hate how there's no observability into subagents and you just get a summary. How do they report back to the orchestrator? Tmux?
257.
by gck1 8mo ago
This looks very similar to what I'm doing. Few questions: - How do you adress spec drift? A new feature can easily affect 2 or 3 specs. Do you update them manually? Is a new feature part of a new spec or you update the spec and then pl
258.
by gck1 8mo ago
> Every decision is traceable to one context window There are no models that can do all the mentioned steps in a single usable context window. This is why subagents or multi-agent orchestrators exist in the first place.
259.
by gck1 8mo ago
You can't sneak in a container and consume more than allocated though. The limits are enforced server side.
260.
by gck1 8mo ago
This is a different kind of shortage though. Previous ones were cyclical and caused by supply/demand mismatches or natural disasters. This one is structural. The manufacturers are actively choosing to prioritize AI because the margins
261.
by gck1 8mo ago
Market is fixing it. Memory makers prioritized HBM and enterprise NAND, some, like Crucial, went out of consumer business entirely. At the same time, the rational market is behaving rationally - they're not increasing production be
262.
by gck1 8mo ago
I'm seeing the same thing. I was consulting a group of people in my city that wanted to digitize massive load of old VHS tapes. No AI, no crazy tech, just standard, boring storage+network infrastructure. I'm looking at the procure
263.
by gck1 8mo ago
Here's an easy experiment to conduct: look around the room at your home and count all the devices that have a CPU, RAM, SSD or HDD. Then take a look at your bank statement to see what are the services you pay for monthly that also requ
264.
by gck1 8mo ago
We see that market is very irrational now and it can stay irrational for long enough to destroy everything we know in tech. By the time market figures things out, you may no longer have services, and hardware that you use daily. When such a
265.
by gck1 8mo ago
No manufacturer is increasing supply though. RAM, SSD, HDD - they just reallocated their existing supply to AI.
266.
by gck1 8mo ago
That, and also because rust compiler is a very good guardrail & feedback mechanism for AI. I made 3 little tools that I use for myself without knowing how to write a single rust line myself.
267.
by gck1 8mo ago
This is likely just the first wave. If this component hoarding by AI continues, and it likely will, at some point, it will be just OpenAI and Anthropic who can afford to have compute. This has affected SSDs first, then RAM, then HDD and it
268.
by gck1 8mo ago
This sounds very plausible. OpenAI has hoarded 40% of world's RAM supply, which they likely have no use for other than to starve competition. They (or other competitors) could be utilizing the same strategy for other hardware. Which is
269.
by gck1 8mo ago
I don't know what it's called when something becomes an irony and then this irony becomes an irony itself, but that's what's up with OpenAI today. On one hand, they started this 'we're closing things down becau
270.
by gck1 8mo ago
Sorry, AI companies have bought all the hard drives for years 2026 and 7. And RAM. And SSDs.

More ›