kolinko

8 ms·

181.
by kolinko 7mo ago
Well no, all the routers on your subnet need to understand it. So let’s say your internet provider owns x.x.x.x, it receives a packet directed to you at x.x.x.x.y.y… , forwards it to your network, but your local router has old software and
182.
by kolinko 7mo ago
I think with zero knowledge proofs we can have it both ways?
183.
by kolinko 7mo ago
Full version or free version? Most conplaints people make, I found, was because they use the free version.
184.
by kolinko 7mo ago
and he’s using a free version of chatgpt? and not publishing source prompts - so there is no way to replicate?
185.
by kolinko 7mo ago
The author uses a free version of ChatGPT. Fails to mention that anywhere, but you can see from the screenshots. And they don’t provide the prompt, so you can’t really verify if a proper model has the same issues.
186.
by kolinko 7mo ago
It also helps that OSX has FreeBSD underneath (so, practically, Linux).
187.
by kolinko 7mo ago
The blog author uses free version of ChatGPT (not logged in on screenshot) - so really talks about the previous generation models. It would be nice if this was mentioned transparently in the beginning of article. I mean - new models also te
188.
by kolinko 7mo ago
Not Claude/Copilot. Claude.
189.
by kolinko 7mo ago
Or formal methods and other tools for verifying the code security?
190.
by kolinko 8mo ago
With Claude you just tell it to set up whatever it needs and you have a smooth access to everything. Mine uses Nanobanana for image generation, Sora for video, Gemini for supplementary image processing and so on. Setting up each one was 5-1
191.
by kolinko 8mo ago
For me it's the opposite - sure, for many outputs I don't need to think, but then I end up thinking on a higher level, and doing even more work. An analogy would be - if GPS allows you to not worry about which turn to take, you ca
192.
by kolinko 8mo ago
I think the author just doesn't know how to use LLMs well. "Because what would be missing isn’t information but the experience. And experience is where intellect actually gets trained." From my experience, LLMs don't cau
193.
by kolinko 8mo ago
you’ve been hearing that since mid 2025 bc that’s when it became true.
194.
by kolinko 8mo ago
Who wants it? On my iPad I want an iPad OS, MacOS is not good for touch operation. MS tried merging two OS-es (surface) and see how it went.
195.
by kolinko 8mo ago
Which version? I just tested mine and it replied with an actual quote
196.
by kolinko 8mo ago
For agent to agent payments, Bitcoin/Ethereum could be quite useful now, as traditional payment protocols and banks are not likely to provide decent APIs any time soon.
197.
by kolinko 8mo ago
Yeah, and the original HTTP had PUT and DELETE methods.
198.
by kolinko 8mo ago
always cc the local GDPR office when reporting such things
199.
by kolinko 8mo ago
Exactly this
200.
by kolinko 8mo ago
Which is how it works with people as well
201.
by kolinko 8mo ago
The new claude/opus, esp, with additional skills is actually pretty decent with UX.
202.
by kolinko 8mo ago
You’re probably getting downvoted because what you said about TCP/IP/HTTP doesn’t make sense.
203.
by kolinko 8mo ago
What parent means is that you can with no problem build over the classic tcp/ip.
204.
by kolinko 8mo ago
I’d love to read more about how the original code looks like - examples of the parts that would be so difficult to transpile / understand and so on. And perhaps an overview of the game’s architecture? It surely is a unique piece of cod
205.
by kolinko 8mo ago
So if something is written by an LLM it makes it an infomercial? Also, what are you comparing this post to? Because you should compare it to the author’s own writing and according to the author, his writing is not that good.
206.
by kolinko 8mo ago
I, for one, enjoyed the read. Would love even more details though!
207.
by kolinko 8mo ago
In my expedience, Claude Code with opus 4.5 is the first one to tackle such issues well.
208.
by kolinko 9mo ago
They didn't test Opus at all, only Sonnet. One of the tasks was "Build an interactive dashboard for exploring data from the World Happiness Report." -- I can't imagine how Opus4.5 could've failed that.
209.
by kolinko 9mo ago
Also >20 years in software. The VSCode/autocomplete, regardless of the model, never worked good for me. But Claude Code is something else - it doesn't do autocomplete per se - it will do modifications, test, if it fails debug,
210.
by kolinko 9mo ago
Are you at all familiar with the architecture of systems like theirs? The reason people don't jump to your conclusion here (and why you get downvoted) is that for anyone familiar with how this is orchestrated on the backend it's o

More ›