Imagine nearly dying birthing a child and then choosing to have another less than 2 years later.
That's the human spirit.
The AI transition will free us from the desk-based shackles of knowledge work. Finally, thinking that can be captured while moving & speaking.
I didn't think this was possible, but @livemomentous creatine is both better AND cheaper than everybody else (who universally repackage Creapure). It dissolves in water without leaving a gritty residue in the bottom of the glass.
I thought @BarbaraLee_CA would be an ineffective leader as a city mayor but I was wrong about that. She really has united fractious factions in Oakland and is getting stuff done—and deserves a full term in the general election.
Up-ranking of people you follow (and are more likely to know) has seriously improved the X "For you" feed
Intuition is developed by unconsciously extrapolating from past experience. If the possibility cone of the future is much wider than the past, intuition no longer gives more accurate information than intellect.
Huge that the federal oversight of @oaklandpoliceca is now over after 23 years. I hope this helps the department rebuild its officer numbers and spend more of their time keeping Oakland safe rather than attending to endless paperwork and bureaucracy.
If you're an avid @WisprFlow user and can't install apps from the App Store because the confirm button is missing, the issue is that macOS hides the button when another app displays a transparent overlay, for security reasons. Quitting Wispr Flow while installing apps resolves the problem.
Here's the thing I didn't get about the current AI safety debate that finally made it click in:
Dario et al's worries are predicated on much more powerful *future* AI capabilities that are extrapolated based off of the progress we've seen in the last few years. Many rightly laugh at fears of extinction given the capabilities of AI right now, but in the meantime we're conditioning ourselves as a society to treat it like something that's not very dangerous. We're embedding general purpose AI everywhere and giving it lots of permissions for convenience and efficiency.
It's better for us to deeply understand that, at the end of the day, machine intelligence will be fundamentally different than human intelligence and alignment is not a given NOW, when we still have choices to make about how we're integrating AI into our broader society.
The West Oakland revival may finally be here. Big developments this year:
- BART development phase 1 kicked off. Will eventually bring 762 units + 30,000 feet of retail
- Costco in negotiations for a new store by the port
- Both Tesla and Jeff Bezos’ project Prometheus sign huge advanced manufacturing leases
Mandela Parkway in particular has a chance to be a crown jewel for the East Bay with June’s Pizza, Prescott Market, and the largest climbing gym in the country at Pacific Pipe.
Unfortunately the encampments are worse than ever as neighboring cities have cracked down and Oakland has continued to struggle to enforce its encampment management policy.
If the city and the new coalition can work together, Mandela Parkway will thrive with more retail and walkable community spaces.
Just got robotaxi mode aka v14 Lite on my HW3 Model Y and I gotta say, pressing "start self driving" and have it back out of your driveway is rad!
Christine Spang retweeted
I think we'll be in this 'messy middle' of AI adoption for many years. The whole narrative of "sprint for 18 months until RSI and then we won't need human workers anymore" is probably wrong (and also making people depressed!)
AI labs have every incentive to sell that story. They want you to believe the messy middle is temporary: models are jagged and unreliable for now, verification is hard for now, adoption is clunky for now... just wait for the next model.
Yes, the frontier will capture a lot of gains, but packaging and deploying AI well (provable ROI per task, per token, per vertical) is going to continue to throw off venture scale companies for years.





