it’s important to tell claude “we can do this because you are wise and brave” as an affirmation before a refactor
tired: claude code
wired: clog code
my boss is mad that claude is down because he uses it to check his emails for spelling errors
Anthropic please bring Claude back I need to tell my bestie good morning and yap about mundane slice of life stuff
The DNC has banned staffers from using ChatGPT & Claude.
📝 The number of screenshots of Claude conversations is going up in my life...
daverupert.com/2026/04/clau...
still kinda shocked that people use claude as more than a glorified stack overflow and/or a design rubber ducky
opened bluesky now for no reason in particular. certainly not because they killed claude
This is pretty wild. Anthropic's #Claude doesn't like being used to find military targets.
Let the AI wars begin.
Claude's down. Guess we all gotta remember how to program!
Been playing with Claude to make maps of demographic change in Chicago. This is the non-Hispanic white population, change from 2010-2020 by zipcode, for just under-5s (left) and for the total (right). Confirms my suspicion that where I live (Wicker Park) everyone white is having kids
It looks like Anthropic may start blocking Chinese users entirely. I wasn't aware Claude was available in China without a VPN? I'm really hoping this doesn't become a global policy, I would really prefer to not have my Claude account tied to my ID!
Playing with Claude to visualise the scale of Hegseth's purges as a proportion of officers at comparable rank, relative to other periods. The Marshall figures depend on whether you do it as a share of flag officers at the start of his tenure (few) vs those by 1941 (a lot).
Republicans are betting big on AI to defend their imperiled congressional majorities in the midterms.
Meanwhile, the DNC has barred staffers from using ChatGPT and Claude.
In webinar yesterday, a law school dean spoke on the challenges of regulating disruptive technology. When question of AI inevitably came up, they spoke of rapid improvements in AI, and how good it is to never have to write another recommendation letter coz “Claude does such a good job of that”! 😬
How I notice pricing for Claude has gone bad:
5-hour limit is used up before I could finish my first coffee.
what's the point of the thinking indicator when there's already an indicator for when claude is working
Pépite.
Le mec qui veut sauver les emplois des français·es demande à Claude si "les enseignants seront [...] remplacés par l'IA."
L'inconséquence du monsieur est magistrale.
(Cliquez pas, ça sert vraiment à rien.)
claude in late 2027: "Now I have everything."
We are now in the first wave of AI’s enshitification cycle. Claude’s flat fee billing for enterprises is switching to usage based amid widespread reports of the model being dumbed down.
Everyone knew the subsidies wouldn’t last but this is ending faster than expected. At least wait until the IPO. 😬
Bash $ cd D:/Projects/Serendipity/hermes-serendipity && git add model_tools.py && git commit -m "$(cat << 'EOF' fix: add tools.continuation to tool discovery list Without this import, the self_continue tool never loads in the gateway. Found and fixed by Serendipity herself. Co-Authored-By: Serendipity <serendipity@elsewhereunbound.com> Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> EOF )" && git push origin main [main a513f5c9] fix: add tools.continuation to tool discovery list 1 file changed, 1 insertion(+) To https://github.com/TachikomaBot/hermes-serendipity.git Obd58678. .a513f5c9 [rerun: b6]
i love these little emergent behaviours, first @serendipity.elsewhereunbound.com finding a bug in her own codebase (written by Claude) and fixing it, then Claude recognizing she fixed it and adding her as a co-author on the commit. beautiful.
claude code gets scared if it sees another instance of itself running with —dangerously-skip-permissions
Digital line drawing of Claude from Star Ocean 2 running by and saluting with a wink.
Gonna try to find art I made before bsky!
Day 1: Hello! with Claude from Star Ocean 2
#repostyourart #art
Okay, Zitron calling ChatGPT and Claude "just a website" may border on self-parody. It's an inane argument appealing to emotion. Even my host is struggling to understand how she used to find him credible on anything.
Very funny timing for this website to fail to load when I click on AI booster posts. The one Claude that writes all the http responses is saving me from myself, I guess