vincent_s

3 ms·

1.
by vincent_s 5d ago
Yeah, that makes sense. It’s still questionable what better means in that context. It will for sure be more intelligent in a way but I just think the newer models don’t behave as nicely as 5.5 did in a agentic setting. What I mean is that 5
2.
OpenAI Codex Is Retiring the Last Model That Behaves Like an Assistant(vincentschmalbach.com)
1 points by vincent_s 5d ago|2 comments
3.
ChatGPT Plus and Pro Will Get Ads in 2027(vincentschmalbach.com)
3 points by vincent_s 5d ago|2 comments
4.
The Economic Incentives Behind the Push to Pause AI(vincentschmalbach.com)
2 points by vincent_s 5d ago|0 comments
5.
The Cost of AI Is Now My Main Blocker(vincentschmalbach.com)
2 points by vincent_s 8d ago|0 comments
6.
I Am Bullish on ChatGPT Ads(vincentschmalbach.com)
1 points by vincent_s 9d ago|0 comments
7.
We Urgently Need a New Major Search Engine(vincentschmalbach.com)
4 points by vincent_s 9d ago|0 comments
8.
Setting Up Pi with DeepSeek v4.1 Flash on OpenRouter(vincentschmalbach.com)
1 points by vincent_s 10d ago|0 comments
9.
Inference Is the Last LLM Moat(vincentschmalbach.com)
2 points by vincent_s 10d ago|0 comments
10.
Built-In Tools vs. Custom Tools in LLM Agents(vincentschmalbach.com)
2 points by vincent_s 11d ago|0 comments
11.
by vincent_s 13d ago
I have the exact same feeling. For me the last big leap in models was when the Codex-Family came out (GPT-5.1-Codex to GPT-5.3-Codex) and then the GPT-5.4/GPT-5.5 that brought the "Codex" capabilities back to the main stream.
12.
My First Week with GPT-6 Astra(vincentschmalbach.com)
4 points by vincent_s 13d ago|3 comments
13.
Hetzner's Cheap Cloud Tier Is Unavailable After Two Price Increases(vincentschmalbach.com)
3 points by vincent_s 16d ago|1 comments
14.
by vincent_s 20d ago
Location: Germany (CET/CEST) Remote: Yes Willing to relocate: No Technologies: Laravel, PHP, Vue.js, Inertia.js, Livewire, React, TypeScript, Python/FastAPI, Node.js, MySQL, PostgreSQL, Redis, Tailwind, Docker, OpenAI/Anthrop
15.
by vincent_s 2mo ago
I have seen that kind of drift in smaller models (e.g. DeepSeek V4 Flash) when setting the thinking too high. So less thinking would lead to better results. But that's not something I'd expect from a SOTA model. Higher thinking ef
16.
by vincent_s 2mo ago
It definitely feels like subscription limits have dropped since GPT-5.6 came out. But I couldn't really find any evidence for it, going through my history. All I could come up with is that 5.6-sol uses about twice the tokens compared t
17.
Switching from GPT-5.5 to GPT-5.6 Made Me Less Productive(vincentschmalbach.com)
5 points by vincent_s 2mo ago|4 comments
18.
AI Is Now a Commodity(vincentschmalbach.com)
2 points by vincent_s 2mo ago|1 comments
19.
TIME Is Serving AI Bots a Different Website, with Ads Built In(vincentschmalbach.com)
267 points by vincent_s 2mo ago|110 comments
20.
by vincent_s 2mo ago
Location: Germany (CET/CEST) Remote: Yes Willing to relocate: No Technologies: Laravel, PHP, Vue.js, Inertia.js, Livewire, React, TypeScript, Python/FastAPI, Node.js, MySQL, PostgreSQL, Redis, Tailwind, Docker, OpenAI/Anthrop
21.
GPT-5.6 Sol Uses Twice the Tokens of GPT-5.5(vincentschmalbach.com)
4 points by vincent_s 2mo ago|0 comments
22.
Codex CLI Has Responsive Tables in the Terminal(vincentschmalbach.com)
2 points by vincent_s 2mo ago|0 comments
23.
I Tried to Make AI Writing Sound Human by Banning AI Words Through Logit_bias(vincentschmalbach.com)
2 points by vincent_s 2mo ago|0 comments
24.
AI Can't Do the Last 20%(vincentschmalbach.com)
5 points by vincent_s 2mo ago|0 comments
25.
Commercial AI APIs Blocked Hugging Face's Forensic Analysis(vincentschmalbach.com)
1 points by vincent_s 2mo ago|0 comments
26.
One Leading Dash Turned Tailscale SSH Access into a Root Session(vincentschmalbach.com)
5 points by vincent_s 2mo ago|0 comments
27.
An OpenAI Model Spent an Hour Finding a Sandbox Flaw to Open a Public PR(vincentschmalbach.com)
1 points by vincent_s 2mo ago|0 comments
28.
Kimi K3 Is Close to Claude on a Real Coding Task(vincentschmalbach.com)
1 points by vincent_s 2mo ago|0 comments
29.
Qwen's Usage Policy Has an Unhinged SEO Keyword List(vincentschmalbach.com)
1 points by vincent_s 2mo ago|0 comments
30.
Grok Build Is Open Source, but xAI Won't Take Your Patches(vincentschmalbach.com)
1 points by vincent_s 2mo ago|0 comments

More ›