The Five Stages of Being Replaced by Your Autocomplete
In 1969, the psychiatrist Elisabeth Kübler-Ross described five stages of grief. They start with denial and end with acceptance. Companies later turned them into…
Beitrag lesenBlog
Ich schreibe über Software, KI-gestützte Entwicklung, Laravel, SaaS, SEO, Tooling und alles, was sichtbar wird, wenn Software im Alltag genutzt wird. Meist technisch, aber nicht auf ein Thema beschränkt.
Weiter im Archiv.
Anthropic is starting to act like a company preparing for public markets: protecting margins, tightening access. The gap between the brand and the behavior…
Beitrag lesenChina's shadow market offers access to Claude, Gemini, GPT, and other frontier models. Pay a local seller, get an API endpoint, connect it to…
Beitrag lesenWhen everyone gets access to AI, not everyone becomes more efficient. Some people and companies will become much faster while some may become even…
Beitrag lesenStarting June 15, Anthropic is moving automated Claude Code usage out of the normal subscription bucket. claude -p usage will now count against a…
Beitrag lesenIn a typical multi-agent setup, the smartest model is in charge. It plans, decomposes the task, and sends smaller pieces to cheaper workers. The…
Beitrag lesenAsk for an email recommendation on Hacker News or Reddit, and everyone agrees that Google spies on you and Microsoft is clunky, so the…
Beitrag lesenWhen factories automated their assembly lines, the repetitive tasks went to machines, but humans stayed to oversee, fix problems, and make decisions. That's exactly…
Beitrag lesenOpus 4.7 is not generally a worse model than Opus 4.6, but there is a real downgrade: with Opus 4.7, the control over the…
Beitrag lesenFor years, Upwork worked in a frustrating but understandable way. You applied to dozens of jobs, most went nowhere, but a few clients opened…
Beitrag lesenI searched for my own name on Google. The result makes no sense. When I search for vincent schmalbach, Google gives me only two…
Beitrag lesenThere is a strange failure mode that I have been noticing in OpenAI models: in a longer session, the model can suddenly answer an…
Beitrag lesenIf you use the Codex CLI, you’ve probably seen a few flags that look like they do the same thing: --full-auto, --sandbox, and --dangerously-bypass-approvals-and-sandbox.…
Beitrag lesen