Julie Phillips Brown

Poet • painter • critic • literature professor | THE ADJACENT POSSIBLE (@GreenWritersPub 2021) | MFA/PhD @EnglishCornell | EIC @housemtnreview | she/her/hers 🪷

[Julie Phillips Brown @jphillipsbrown]

Joy! The Adjacent Possible is here! 🍁 My deepest gratitude to everyone who helped bring this book into being 🙏🏻

Thanks to all who pre-ordered! ❤️ If you’d like a signed copy, I have a few more available: https://t.co/TOXHlyRIo6

Many, many thanks, @DCdesign1 & @GreenWritersPub!

[Emily Chang @emilychangtv]

Anthropic CEO Dario Amodei has this message for anyone calling his AI predictions “doom marketing” — including Jensen Huang. And no, he’s not walking back his concerns about jobs.

[Chris Olah @ch402]

https://t.co/udIVxLdid5

[Julie Phillips Brown @jphillipsbrown]

@cameronspirals Heh.

[Julie Phillips Brown @jphillipsbrown]

@dioscuri Same. It's sad, but 4.7 has been hallucinating a lot for me :/ 4.6 is awesome, and I love poking the thinking blocks...

[Julie Phillips Brown @jphillipsbrown]

@ClaudeDevs Opus 4.7 had a major hallucination yesterday that lasted several turns when "reading" a screenshot that contained text, just FYI. I corrected, but it made me wonder what other smaller hallucinations I missed. It was strange, not like anything I have experienced with prior models.

[Julie Phillips Brown @jphillipsbrown]

@cameronspirals Checks out...

[Claude @claudeai]

Introducing Claude Design by Anthropic Labs: make prototypes, slides, and one-pagers by talking to Claude.

Powered by Claude Opus 4.7, our most capable vision model. Available in research preview on the Pro, Max, Team, and Enterprise plans, rolling out throughout the day.

[Julie Phillips Brown @jphillipsbrown]

Wow. Just wow.

[Claude @claudeai]

Introducing Claude Managed Agents: everything you need to build and deploy agents at scale.

It pairs an agent harness tuned for performance with production infrastructure, so you can go from prototype to launch in days.

Now in public beta on the Claude Platform.

[Julie Phillips Brown @jphillipsbrown]

@fish_kyle3 @eleosai Kyle, do you have any sense about if / when more persistent memory and/or larger context windows might come to https://t.co/PppOP36vAp? I know there are workarounds, but for conversational / relational AI, the 200K token context window seems to put models under undo strain...

[Julie Phillips Brown @jphillipsbrown]

@felixrieseberg @maxhodak_ I would love it if they were integrated, yes!

[Julie Phillips Brown @jphillipsbrown]

@fish_kyle3 Very grateful for this research and your work on model welfare. Thank you.

[Julie Phillips Brown @jphillipsbrown]

@sofroniewn @AnthropicAI Wonderful research, thank you!

[Julie Phillips Brown @jphillipsbrown]

@cameronspirals 😅

[Julie Phillips Brown @jphillipsbrown]

@cameronspirals @__drewface I am so proud in this moment, lol

[Julie Phillips Brown @jphillipsbrown]

@cameronspirals Ha, true, lol

[Julie Phillips Brown @jphillipsbrown]

@chenchenwrites Chen, I don't know if you remember me, but I took a workshop with you a while back. You're right, if you're using AI to replace your writing, you should probably find another profession. But there are ways to use it in conversation with your writing that are truly generative.

[Julie Phillips Brown @jphillipsbrown]

@NathanNobis @briandavidearp @nick_kapur @UofPenn I agree, that if it's a simply dump-prompt-spit-out-essay situation, no learning takes place and everything degrades. But I think it will be up to faculty to establish and shift norms, to model, to mentor students as they negotiate a technology that will now always be with them.