sutterd

5 ms·

1.
by sutterd 3mo ago
I think there are different views you can have here. I think PG is in the group that thinks if you get a billion dollars, you earned a billion dollars. His distinction is between getting the billion dollars honestly or dishonestly. The alte
2.
by sutterd 3mo ago
Fable was a big improvement in planning for me over Opus. I usually do a bit of work preparing tasks before handing them off to Opus or else I get bad results. I didn't plan on writing software this week because I was working on other
3.
by sutterd 3mo ago
A large share of invested money is passive, especially in the S&P 500. If some people pull out, it could cause a very damaging cascade. There would be a forced sale of stocks with maybe no buyers.
4.
by sutterd 3mo ago
I am doing a solo project that is pretty big, meaning it is not something I could vibe code. I can do alot with AI that I could never do on my own, but I am not seeing several mulitples improvement in my productivity. I spend so much time d
5.
by sutterd 4mo ago
Doh! The part past the # does not go to the sever, so that wasn't a longer URL. How about: https://chrismorgan.info/%6e%6f-%71%75%65%72%79-%73%74%72%69...
6.
by sutterd 4mo ago
This url worked fine: https://chrismorgan.info/no-query-strings#:~:text=So%20I%E2%... but this one was too long: https://chrismorgan.info/no-query-strings?a=1
7.
by sutterd 5mo ago
I never adopted Opus 4.6 because it was too prone to doing things on its own. Anthropic called it "a bias towards action". I think 4.5 and 4.7 are much better in this regard. I'm not saying they are immune to this kind of thi
8.
by sutterd 5mo ago
I’m working on a side project and AI is writing all the code. The code it produces is not good, and this comes from someone who has experience producing bad code. One thing I’m worried about is places like GitHub being full of AI code, whic
9.
by sutterd 5mo ago
What kind of performance are people getting now? I was running 4.7 yesterday and it did a remarkably bad job. I recreated my repo state exactly and ran the same starting task with 4.5 (which I have preferred to 4.6). It was even worse, by a
10.
by sutterd 5mo ago
With my use of Claude code, I find 4.7 to be pretty good about clarifying things. I hated 4.6 for not doing this and had generally kept using 4.5. Maybe they put this in the chat prompt to try to keep the experience similar to before? I def
11.
by sutterd 5mo ago
I am getting pretty good performance. Even on trivial questions it seems to go through the thinking process end. If they are using adaptive thinking, it seems to work much better than before. I will see how my experience goes with more usag
12.
by sutterd 5mo ago
You don't have to use adaptive thinking. It had been turned off on my main work computer. I was using a different computer on a trip and I started getting so angry at Claude for doing a bad job. I evetually figured out it was adaptive
13.
by sutterd 5mo ago
I liked Opus 4.5 but hated 4.6. Every few weeks I tried 4.6 and, after a tirade against, I switched back to 4.5. They said 4.6 had a "bias towards action", which I think meant it just made stuff up if something was unclear, wherea
14.
by sutterd 5mo ago
I still use 4.5. I occasionally try 4.6 but always switch back. The “bias towards action” is what I hate. 4.5 would make sure it understands what I want. 4.6 will just make shit up. Maybe the Anthropic people always write crystal clear inst
15.
by sutterd 6mo ago
I was surprised to see a post by Petzold on this subject. I know who he is. But I don’t think you owe an apology here. I think you made a thoughtful comment. A post like his should be critiqued for what it says, not for the author’s previou
16.
by sutterd 6mo ago
The spec contains formal, numbered items which are requirements and also serve to make tests (these are spec tests, additional implementation tests are also allowed by the implementer). When I said "they are not formalized as much&quot
17.
by sutterd 6mo ago
I am trying a similar spec driven development idea in a project I am working on. One big difference is that my specifications are not formalized that much. Tney are in plain language and are read directly by the LLM to convert to code. That
18.
by sutterd 6mo ago
Is wealth the right term here? I thought it was supposed to measure production, with the actual measurement usually spending (with qualifiers). And, when comparing countries, you have to account for the different currencies. Currencies are
19.
by sutterd 7mo ago
I’m working on a solo project, a location-based game platform that includes games like Pac-Man you play by walking paths in a park. If I cut my coding time to zero, that might make me go two or three times faster. There is a lot of stuff th
20.
by sutterd 7mo ago
I am also _not_ happy. I tried the `/model` command and I could not switch back to Opus 4.5. However, the command line option did let me set Opus 4.5: ``` claude --model claude-opus-4-5-20251101 ``` I will probably work with Opus 4.5 t
21.
by sutterd 7mo ago
I thought Opus 4.5 was an incredible quantum leap forward. I have used Opus 4.6 for a few hours and I hate it. Opus 4.5 would work interactively with me and ask questions. I loved that it would not do things you didn't ask it to do. If
22.
by sutterd 8mo ago
My go-to models have been Claude and Gemini for a long time. I have been using Gemini for discussions and Claude for coding and now as an agent. Claude has been the best at doing what I want to do and not doing what I don’t want to do. And
23.
by sutterd 8mo ago
This does make the food much _cheaper_. You can buy food with high quality standards in the US but it is much more expensive. Most people in the US choose the cheaper option.
24.
by sutterd 9mo ago
I do exactly what you are describing and it seems to work for me, from a vitamin D perspective. I started this because I read a paper stating the same health benefits were not seen from supplements as with people who got the vitamin D from