🧠Mind expanding links 4 May 2025: consciousness alone truly exists🤯 OMG claude code🚀 brokenism🔨 capital beyond ROI💰 inside arXiv📚

2. Yes, and don't ask again this session

Hi,

Hope you’re enjoying your weekend. Here’s my latest, *intermittent* 🧠Mind Expanding Links roundup of longer reads and deeper thinks which have been keeping my overactive mind occupied recently.¹ (Outside of compulsively scanning through the AI and technology news wires, natch...)

The usual eclectic mix in this issue, hopefully something of interest for everyone:

• 💪On AI, privacy and power

  • 🤯Consciousness alone truly exists

  • 🚀OMG Claude Code

  • 👁️‍🗨️Why I support privacy

  • 👑Oligarchy, State, and Cryptopia

  • 🏰Resisting Digital Feudalism

  • 🚁Let Them Eat Drones

• 🔧On “skilled” labour

  • 🪜Nested hierarchies of skills drive career advancement

  • 🗺️Geography of GenAI labour impacts

• 🧠On brains

  • 🧭Brain mapping reveals similar neural patterns in real and imagined navigation

  • 📱Smartphone break for 72 hours alters brain activity patterns

  • 📉Brain decline begins at 44

  • 🍺🍺Two beers daily ages your brain a decade

• 🏛️On politics

  • 🔨Brokenism

  • 🔥Burning Down The House

  • 💰Capital beyond ROI

• 🧩Eclectica

  • 📚Inside arXiv

  • 🥗Metabolism, but not as you know it

  • 🌌Dark energy may reverse universe expansion

(Reminder: paid subscribers can see these posts in development, plus AI-generated analysis of the stories contained in the Memia Sensorium app.)

---

💪On AI, privacy and power

🤯Consciousness alone truly exists

What GPT 4.5 distils about consciousness, via Sam Altman. (Take with a pinch of salt… lots of pre-training and fine tuning in here to swing this one way rather than the other… but a glimpse of something bigger, nonetheless.)

@sama

🚀OMG Claude Code

I’ve been using Cursor for over 8 months now as my AI-enhanced IDE of choice, but for whatever reason their payments provider still won’t accept my credit card (I’m not alone… take my money already!)

So this week after I ran out of free credits again, I tried out the new Claude Code agentic AI coding app from Anthropic. Here’s my end-of-Friday-afternoon LinkedIn post after completing what I estimated would be a week’s worth of AI-assisted coding in… 3.5 hours:

Incredible. Another order-of-magnitude coding productivity boost *again* (but not free… chews through Anthropic API tokens… but when you think of the acceleration it provides, the value-for-money is exceptional).

The minimalistic command-line UI experience takes a few moments to get used to… but thereafter coding becomes a procession of pressing “1” over and over again, with near-perfect first-shot results each time:

(So far I haven’t dared press “2” … “Yes, and don’t ask again this session” … how crazy is that to just let the AI agent loose with a macro goal and it just achieves what we asked it to do… all current cognitive work is going to head this way far sooner than most people are anticipating…)

👁️‍🗨️Why I support privacy

Vitalik Buterin makes a compelling case for privacy in the digital age, as tech giants use AI to gain unprecedented access to our minds: