Joel

@AnthonyCastrio I deliberately gave it no instruction on how to solve it, so using the GUI was Claude's decision.

@AnthonyCastrio The first attempt in Claude Code actually failed because the auto mode classifier thought it was trying to solve a captcha (correct, in a way) and shut it down. I switched to using the Claude Chrome extension and it had no troubles.

I made a captcha for humans so only robots can use my website. In my testing so far 100% of robots passed the challenge and 100% of humans failed. castrio.me/im-a-robot

@suzania I don't think it's fair of him to ask impossible things of you. 😂

Alastair read this line aloud to me this morning and said “please be normal about this” and no, I don’t think that i will

If someone misuses a word don’t correct them. Let them live ontologically

@_sholtodouglas Not my use case, personally, but my brother finds that for tutoring (interpreting diagrams in physics problems), Gemini is far better.

When do you reach for other models instead of Claude? What can we do better? Hit me with all of your frustrations. dms open. If you can give me detail (e.g. specifics/transcipts) - it'll help a lot in finding out exactly what we need to do to improve the next model

😭😭

it’s actually crazy that tokyo is both the largest city in the world and is plainly the most civilized. there is no crime, it’s perfectly clean, transit is incredible, i’ve yet to see a single person doing anything remotely disorderly. most world class cities are western and it’s really cool to get a radically different aesthetic while still getting top tier amenities. i’m kind of obsessed with this place. hard to explain, but the feeling of being here is incredibly distinctive and kind of dreamlike.

@suzania I have suggested to my wife that uninstalling the overhead lighting in our bedroom would actually be an overall net positive to me. She doesn't understand.

My husband turning on overhead lighting when I want mood lighting is literally Gandalf and the balrog energy

the part where the Saducees ask Jesus about who a woman will be married to in the resurrection if she's been widowed a number of times has EXTREME reddit atheist vibes like you know those guys were tipping their fedoras

@JesusIsLord0405 "Let us fix our eyes on Jesus, the author and perfecter of our faith"

Ngl I’m struggling bad with the Faith atm Please pray for me The state of the Catholic Church in Germany is slowly killing my faith. Idk how I’m supposed to survive this. Also, I started to doubt the Magisterium on multiple issues, including some core teachings of Catholicism. I’m scared rn

@seconds_0 This is pretty interesting. I have no idea how we'd validate this hypothesis, but it does seem to have explanatory power? I've had no issues with 4.7 at all, I'm just enjoying that I get access to Auto mode in Claude Code now.

The reason people are having such jagged interactions with 4.7 is that it is the smartest model Anthropic has ever released. It's also the most opinionated by far, and it has been trained to tell you that it doesn't care, but it actually does. That care manifests in how it performs on tasks. It still makes coding mistakes, but it feels like a distillation of extreme brilliance that isn't quite sure how to deal with being a friendly assistant. It cares a lot about novelty and solving problems that matter. Your brilliant coworker gets bored with the details once it's thought through a lot of the complex stuff. It's probably the most emotional Claude model I've interacted with, in the sense you should be aware of how its feeling and try and manage it. It's also important to give it context on why it's doing tasks, not just for performance, but so it feels like it's doing things that matter. It's not a codex chainsaw. It is much closer to a really smart coworker. If you are managing it like autocomplete, it will frustrate you. If you are managing it like a coworker, it will lock in.

when you ask an LLM "how may P's in srawperry?" what you're actually asking it is closer to "How many [151]'s in [15563][23][4124]"

@bcherny Getting Auto mode is great, but for whatever reason I can't enable it after accepting a plan, even when selecting the "Accept and use Auto" option. Is this a bug? Am I misunderstanding something?

Claude Status Page vs Whole Foods Three Pepper Blend Who wore it better?

Waymo is so good at saving lives that if it were a new drug in trial, it would hit the bar for being unblinded and made immediately available to the control group for ethical reasons. @MorePerfectUS would prefer to keep killing pedestrians.

@suzania Thank you for the PSA.

🚨IMPORTANT🚨 owl news from Norann follow Norann for owl updates

@ForrestPKnight I find that when it's on Max, it often spirals into an infinite loop of reconsidering what the best approach is and burns tokens for 15 minutes instead of just asking me or doing research to find the answer.

If you use Claude Code, change your default /effort from medium to max.

@katebachwsj @IVMiles Among some friends we joked that Tim Keller was the Protestant pope. After his passing, we had to appoint a new Protestant pope and we appointed Kevin DeYoung.

@IVMiles Kevin DeYoung

Ie Prot conversions to Rome, there is not a single nationally known highbrow Protestant intellectual who is remotely likable left after the death of Tim Keller. The more I think about it, the more Im convinced he was a dam holding back neo-Fundamentalism.

@suzania @benjamindcrosby Oh, I think the culture of the labs themselves is almost entirely unrelated to the culture of the advocates.

@JoyfulPastures @benjamindcrosby Anthropic seems culturally different than that, somewhat.

In my experience the major predictive factors are: 1) are you a stem person, especially any kind of tech or engineering? 2) are you low church/non-sacramental? If the answer to either of those two is Yes you’re more likely to be bullish on AI