Nick Young

A paper by Enrico Terrone and me has just been published online in the Philosophical Quarterly: academic.oup.com/.../10.1093/pq…... It's called 'Generative AI and the Medium of Gardening.' In it we say that using @midjourney is like growing a garden. 1/2

@simonw I'm pretty sure you can't. That's one of the reasons for all the angry tweets last week.

Claude Opus 4.7 with adaptive thinking via the API... am I missing something or is it not possible any more to force it to think? (Prompt hacks like "think step by step" don't count here, I mean the equivalent of budget_tokens or effort: high in previous Claude models)

@sean_t_strong @emollick No one wants it tuned. Just give us the option like we had in other versions.

Hey Ethan! Sean here, PM on Claude.ai - thanks for the feedback. This isn't a router, this is the model being trained to decide when to think based on the context -- we've been running this for a while in Sonnet 4.6 in Claude.ai as well as Claude Code. Understood that it's not tuned perfectly in claude.ai yet - we're sprinting on tuning this more internally and should have some updates here shortly. Feel free to DM us examples of queries where you expected thinking and didn't see it

@bcherny @mark_k @AnthropicAI yeah, it has so far decided not to bother thinking one single time. Even when i tell it to, or use 'ultrathink' or whatever, it doesn't do anything. I shouldn't have to be wasting my time persuading it to think anyway. It's useless to me now.

@mark_k @AnthropicAI Not accurate. Adaptive thinking lets the model decide when to think, which performs better. Opus 4.7 also uses more thinking tokens on average than 4.6, which is why we have increased rate limits for all subscribers to make up for it.

Seems like Opus 4.7 is mostly a compute-saving release. The new 'Adaptive Thinking' setting basically keeps Thinking disabled by default, which nerfs performance but significantly cuts inference GPU usage. @AnthropicAI is clearly struggling with capacity, so it makes sense.

@theo Hate it. adaptive thinking is dreadful, makes it more or less unusable for my use case.

How are people feeling about opus 4.7 so far?

@emollick Opus 4.6 was my favourite model for analytic philosophy, all ruined now because 4.7 doesn't recognise philosophical problems as worth thinking about.

I think the adaptive thinking requirement in Claude Opus 4.7 is bad in the ways that all AI effort routers are bad, but magnified by the fact that there is no manual override like in ChatGPT. It regularly decides that non-math/code stuff is "low effort" & produces worse results.

@bcherny @realsigridjin it really isn't better. It has ruined the way I use opus. Just cancelled my max subscription.

@realsigridjin Better to let the model think. It's sort of like how we used to manually set "temperature" two years ago -- nowadays it's better to let the model decide.

in claude web - opus 4.7: only adaptive thinking mode - opus 4.6: i can turn on/off reasoning mode so basically you can't control thinking mode

@claudeai @AnthropicAI But how long until you make it dumber to cut costs?

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.

@kepano I've just been playing around with reader and is *so* nice.

New in Obsidian Reader: - Themes and typography settings - Highlighting - Save options Also added a beautiful new reading experience for deeply nested comments on Reddit and Hacker News. Available with Obsidian Web Clipper 1.3 on all browsers.

@kinginmotion Inside with the Agent Client plugin running claude and codex.

Claude Code + Obsidian users — where are you running it from? Inside Obsidian or outside?

@claudeai if I could edit the claude.md file like i can in Claude Code, then i'd use Cowork all the time. (also, could you please let me set my own keyboard shortcuts in the Claude desktop app please?)

Claude Cowork is now available to Pro subscribers.

@felixrieseberg I spent the last week getting claude code set up with my obsidian notes, I tried out cowork last night and it seems promising but much more locked down than claude code. for example, I cannot edit the claude.md file itself.

👋 Hi, I'm Felix and I work on Claude Cowork, bringing Claude Code closer to all kinds of knowledge work. It's an early and rough preview, please tag me in any feedback - we want to iterate very quickly and make it a little better every day.

@FlixBus_IT for anyone else who wants to complain or get a refund the form is here. docs.google.com/forms/d/e/1FAI…

@FlixBus_IT you just cancelled my trip and didn't tell me where I can ask for a refund. This is bad and dishonest customer service. Can you redeem yourselves here by telling me who I should contact?

@peduarte @raycast i'm another vote for more advanced ai based stuff. I know how to use ai commands, ai extensions, MCP etc., but i think i am probably missing out on how to get the best out of them

gonna be making a few new @raycast videos next week what do you want us to talk about? tell me what videos you want so we don't need to ready your mind. thank you very much!

@emollick Same prompt. It chose "The Observer Effect Society - Humans discover they're quantum beings who only exist when observed, leading to mandatory 24/7 mutual surveillance for survival." Like yours, there is a screaming wife.

Claude is pretty funny: “Give me 10 brilliant ideas for a science fiction short short story, pick the most brilliant and execute it terribly” It picked: “People start receiving Amazon packages from parallel universes where they made different life choices.” And it was terrible

@arcinternet @browsercompany Don't believe you. You ditched a good product for💵💲💵💲, which is understandable but a lesson leaned for your users. Also Dia is terrible.

You may have heard the news! @browsercompany has entered into an agreement to be acquired by $TEAM. Here’s what it means for our Arc Members like you: - Arc and Arc Search will continue to exist - Same team, same way of building - We’re growing Dia, our AI browser

@thealexbanks was this not published at launch, or has it been updated?

OpenAI just published their official prompting guide for GPT-5. Master these 6 critical prompting techniques:

@iceGearMusic Mersenne is one of my favourite ever synths, and I have all of your others. Happy to hear you'll be making more.

Thank you very much for all the kind comments. I truly appreciate your support. I will continue to create products that focus on the joy of shaping sound by hand, staying true to the approach I believe in.

Some may find my apps difficult or miss having more presets. I approach my work more as an engineer than a musician, creating tools for those who enjoy shaping sound themselves. If they do not suit you, please consider requesting a refund: support.apple.com/HT204084

@peduarte I like the 'glanceable' year in progress extension –i have it as a favourite. I always thought it would be nice to have one for the weather. Nothing fancy. a cloud/sun/etc., maximum temperature..

whats a raycast extension that you wish existed?

@elder_plinius The Waldo Moment episode of Black Mirror

What’s the most prophetic movie?