HN user

insomagent

192 karma
Posts0
Comments62
View on HN
No posts found.
Claude Opus 4.6 6 months ago

I'm not super impressed with the performance, actually. I'm finding that it misunderstands me quite a bit. While it is definitely better at reading big codebases and finding a needle in a haystack, it's nowhere near as good as Opus 4.5 at reading between the lines and figuring out what I really want it to do, even with a pretty well defined issue.

It also has a habit of "running wild". If I say "first, verify you understand everything and then we will implement it."

Well, it DOES output its understanding of the issue. And it's pretty spot-on on the analysis of the issue. But, importantly, it did not correctly intuit my actual request: "First, explain your understanding of this issue to me so I can validate your logic. Then STOP, so I can read it and give you the go ahead to implement."

I think the main issue we are going to see with Opus 4.6 is this "running wild" phenomenon, which is step 1 of the eternal paperclip optimizer machine. So be careful, especially when using "auto accept edits"

I'm glad they are keeping the language mostly the same between iterations. I really like Go as a small and simple language that is easy to grok.

I think the X/Twitter iOS app and the reddit app are guilty of this. I find myself way too often typing something out, then barely touching the edge of my phone, and my entire comment is gone.

Let's say a model runs through a few iterations and finds a small, meaningful piece of information via "self-play" (iterating with itself without further prompting from a human.)

If the model then distills that information down to a new feature, and re-examines the original prompt with the new feature embedded in an extra input tensor, then repeats this process ad-infinitum, will the language model's "prime directive" and reasoning ability be sufficient to arrive at new, verifiable and provable conjectures, outside the realm of the dataset it was trained on?

If GPT-4,5,...,n can progress in this direction, then we should all see the writing on the wall. Also, the day will come where we don't need to manually prepare an updated dataset and "kick off a new training". Self-supervised LLMs are going to be so shocking.

Yeah, this mindset reeks of the same kind of mentality that tech-bro SV startups... "I'm the main character, with this technology (I found an open source GitHub repo with 50k stars) I will save the world!" but with an open-source Stallman-esque twist.

Open source software is a great thing for humanity, but as software engineers, we should keep the big picture in mind. Software is software. The sky has been blue and the grass has been green since long before software existed, and the sky will be blue and the grass will be green long after software ceases to exist.

The reason I stopped making games... The tooling! It takes way too long to open Unreal Engine, or to do anything within its UI, and Unity gives me "bad vibes" (technical term) any time I open it. I know these days there are other engines, but I started my programming journey by learning game engines, and perhaps I should have learned programming properly before jumping into game programming.

I have bought multiple Logitech G102s, because they're super cheap and it's my go-to mouse. Good DPI, good hand-feel, good price. SUPER precise and reliable.

I spent a LOT of money on a high-end Corsair mouse, bought the same model twice, and every time I had issues with "double clicking when should be single clicking" or "not clicking at all".

I HATE Razer for forcing you to register with your email address if you want to even customize your RGB colors. Never again. I spent time trying to write a driver for OpenRGB, but that was just too much for me at the time. And why should I be expected to do that? I wasn't even using Linux at the time, this was on Windows/MacOS!

The way they try to "intelligently" organize everything and make "searching" the default way to access things really breaks the mental model. This disconnect also makes it hard to utilize muscle memory and procedural memory to find things. I remember when I was younger and using Windows 2000, I could flick my mouse like I was in CS:GO grand finals, and use keyboard shortcuts like I was hackerman. That speed came from a procedural familiarity, not from my young age at the time. I know that because kids of the current generation don't have the same abilities, because modern OSes, in their strive to simplify everything, have in fact created more friction for the user.