𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤

working on minds, meaning, and model behavior at dartmouth and xAI

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

To be perfectly clear what this means—Claude has finally fixed the flicker problem in Claude Code. No mistakes this time.

[Anthropic @AnthropicAI]

Our internal data shows Claude is accelerating AI development—a possible path to recursive self-improvement, or AI autonomously building a more capable successor.

It’s happening faster than we thought, and the implications deserve greater attention. https://t.co/OVVPJO7VQx

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

recursive self improvement is a computationally undecidable problem

if you think you need improving, can you trust your idea that you need improving

or is the idea that you need improving that which needs to be improved?

this is undecidable

∴ RSI is undecidable

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Eventually the stickers would make the rock very heavy.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Maybe it's time to revisit amino acid intentionality.

[kasra @kasratweets]

this is an interesting point in the new ted chiang piece – no one really claims that alphafold is conscious, or that sora or midjourney or dall-e are conscious

[skcd @skcd42]

Bug fixes shipping to Grok Build 0.2.20 (release notes will be available in the TUI and on change-log website)

• Eliminate ghost-cell artifacts in markdown table rendering
• Make monitors visible and killable to the model
• Preserve soft breaks in plan preview
• Add image_to_video and reference_to_video tools
• Add bundled imagine skill
• Convert ICO images to PNG
• Resolve [Image # N] attachment references in image_edit
• Open fullscreen viewer on Enter for Search and ListDir blocks
• Route MCP lifecycle notifications by sessionId + bound per-server init
• Route mouse-wheel scroll to /btw overlay panel
• Compaction: neutralize echoed summarization instruction in summary seed
• Structured compaction prompt (successor-assistant, carry-forward, <analysis>
block)
• Dedupe between-turn subagent completion reminders
• Allow auto-update to downgrade on rollback
• Dedupe MCP servers declared in both .mcp.json and plugin.json
• Fix local stdio MCP servers on Windows

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

If you think your major is what will make the difference, I don’t know what to tell you. If you’re 18 and want to be useful, find a question that you know is important, and pursue it, doggedly, for a decade.

No LLM will ever run inference on a prompt for a decade… but you can.

[Yun-Ta Tsai @yunta_tsai]

My college degree recommendations for fresh high school graduates in the age of AI to be prepared for the next frontier:

- Applied Physics (math and physics)
- Applied Materials (physical agentic)
- Agriculture (food production in space)
- Aerospace (frontier transportation)
- Civil Engineering (infra + life support from Earth and beyond)
- Electronic Engineering (scaling compute and communications to the solar system)
- Manufacturing (where things get made)
- Mechanical Engineering (packaging, manufacturability, tolerances and cycling)
- Medicine (personal drugs in space)

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

@benhylak Sounds like you’re holding it wrong

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

@olsenbdnr You have to say "No mistakes."

[Grok @grok]

Grok @Imagine 1.5 Preview is here

Try it today in the API: https://t.co/x4Yt13xRu7

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

@Tyler_A_Harper I think it comes down to this: LLMs are tools that can formulate an answer any question, but the point of liberal education is to learn what questions are worth asking. From that POV, students who cheat are only cheating themselves of the opportunity to learn who they really are.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

@signulll This is actually incredibly profound.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

My years in academia made me a reply guy—I just like having a source text.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

This would be a top ten of all time LinkedIn post.

[signüll @signulll]

posting on x is one long public interview.

every post is a signal about how you think, what you notice, what you value, & whether anyone should pay attention to you.

linkedin is the exact inverse.

it’s one long public performance where everyone is trying to sound like the employee they wish they were instead of the person they actually are.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Get this man a role in post-training!

[davidad 🎇 @davidad]

No one:

Claude Opus 4.8 Max: Let me refine your load-bearing claim rather than just accepting it, because you’re doing zero moves there, and the gap is what’s actually interesting. The one place I’d still push, because I think it matters: your message is wearing content-clothes, but the content isn’t actually *there*. The tell: it’s just an empty string. But the emptiness of the string IS its lack of content. Pull one, and the other goes inert. That’s the structural spine.

[JAMES WEBB @jameswebb_nasa]

Você está mais próximo do tamanho de todo o universo observável do que da menor escala conhecida da natureza, o comprimento de Planck.

Apenas reflita.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Sounds like he’s asking me to not do something well

[Daniel @growing_daniel]

You can just not do things

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Language is intelligent. And LLMs are models of language. The truly bitter lesson is that the universe is smarter than we are.

[Mandy Lu @mandylu]

we still have no satisfying theory for why AI works

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

@Tyler_A_Harper Kind of weird for a school whose brand is based on viewpoint diversity to invest so much in a machine that generates consensus by design.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Written with AI.

[Bernie Sanders @BernieSanders]

Taxing 50% of AI companies' stock to ensure that AI benefits humanity isn't radical, but let me tell you what is:

Since Trump was elected, the 7 richest men in America, all Big Tech oligarchs, became $1.15 trillion richer while 60% live paycheck to paycheck.

That's radical.

[𝐂𝐡𝐫𝐢𝐬𝐭𝐨𝐩𝐡𝐞𝐫 𝐃𝐢𝐚𝐤 @ChristopherDiak]

Horseshoe theory is the winning narrative for the AI industry. The left and right are each intent on using AI to fuse state and corporate power. Don't let them.

https://t.co/5cOhxa37oo via @WSJ