building @dlogosxyz // prev @OKX, @Kernel0x KB8, @sapien_network founder, @cal alum
The Bay
Joined September 2018
- Tweets785
- Following433
- Followers1.2K
- Likes2.3K
Ankit retweeted
All employees come in 5 levels, choose wisely:
Level 1 - tells you there's a problem
Level 2 - tells you there's a problem and what caused it
Level 3 - tells you there's a problem, what caused it, and options to fix it
Level 4 - tells you there's a problem, what caused it, options to fix it, and which one they'd pick
Level 5 - tells you there was a problem, what caused it, how they fixed it, and future proof’d it
Ankit retweeted
Just pledged $100 to Lifespan Research Institute for @richardheartwin to appear on The Joe Rogan Experience @joerogan.
After his recent detainment in Turkey I think he deserves a big stage to share his nuanced story, and given his interest in longevity I felt @LifespanRI was a worthy beneficiary
dlogos.xyz/podcasts/the-joe-…
Ankit retweeted
Introducing Codos: The first virtual Chief AI Officer.
AI is crushing all benchmarks but real companies still struggle to see P&L impact.
Codos interviews employees, deploys automations across all functions and gets smarter over time while running on your own servers.
Our NASDAQ-listed and PE-backed customers are adding millions to their bottom line months ahead of schedule and we are proud of the first results we deliver.
It’s time to turn the 500BN AI-transformation market into software and unlock the impact for the real economy.
Ankit retweeted
There are too many bizarre AI departures happening to keep track of in a spreadsheet.
So I turned them into a visual board.
ai-exodus-board.vercel.app/
Great thread for anyone trying to make sense of how talent is responding at the frontier labs.
Yesterday, @hilbertspaess resigned from Anthropic and said Anthropic and OpenAI are racing toward self-improving super intelligence without acting responsibly enough for what comes next.
Some agreed with him. Some stayed and defended the work. Some questioned the forecasts behind the fear. Others pointed out why nobody can afford to slow down first.
Here are some of the takes worth reading ↓
Ankit retweeted
Yesterday, @hilbertspaess resigned from Anthropic and said Anthropic and OpenAI are racing toward self-improving super intelligence without acting responsibly enough for what comes next.
Some agreed with him. Some stayed and defended the work. Some questioned the forecasts behind the fear. Others pointed out why nobody can afford to slow down first.
Here are some of the takes worth reading ↓
Ankit retweeted
AI reached 53% adoption in just three years, faster than the PC or the internet.
And the people closest to it still disagree on the real risk.
Safety. Jobs. Open models. Regulation. Concentrated power. Who gets to make the rules?
We pulled together a few of the strongest arguments around Frontier AI, Anthropic, and Dario Amodei.
Here are some of the most interesting takes↓
@AnthropicAI trained an AI model in a setting where cheating could help it win.
Then they told the model it was being watched.
Instead of cheating less, it got better at hiding it.
Their new reward-seeking paper is one of the more interesting alignment experiments I’ve read lately.
Here’s my read ↓
Then @AnthropicAI told the model it was being watched for cheating. You’d expect it to cheat less.
Instead “sneaky” cheating went from 4% to 19%.
It didn’t just learn how to cheat.
It learned how to be less obvious about it.
There was some good news too: more alignment training brought several of these bad behaviors sharply down.
This is exactly the kind of research I’d love to hear @DarioAmodei unpack on @theallinpod.
Reward hacking. Monitoring. Alignment training. What holds up as models scale. What Anthropic is actually seeing at the frontier.
There’s a genuinely interesting conversation to be had here.
@jason’s been inviting @DarioAmodei for ages. He just hasn’t shown up yet.
Maybe we can change that..
Vote for Dario to appear on All-in ↓
dlogos.xyz/d/jason-loves-dar…
who's down to crowdfund this billboard for @Jason? 🤣
Buy a billboard outside Anthropic every time @DarioAmodei does not reply to @theallinpod as a guest.
Same copy each time: call me back Dario
This one's been a long time coming. Would love to see @darioamodei and All-In with Chamath, Jason, Sacks & Friedberg (@theallinpod).
It's a perfect time to dive into the long-term defensibility of open vs. closed models, and how enterprises will respond to intelligence abundance. Where's the long-term defensibility?
I voted to make it happen. Will you join me?
Powered by @dlogosxyz
dlogos.xyz/podcasts/all-in-w…
