Paige

First principle & Longtermism | building something..

Open Robotics @FabricFND founding team @openmind_agi , ex @McKinsey

[Paige @paigeinsf]

https://t.co/4tB4HdywbD

[Paige @paigeinsf]

@YorkYang5050 Thanks for sharing!

[Paige @paigeinsf]

Really good piece from someone who’s actually been in the trenches through the AV era.

Not every robotics problem has the same economic value.

E.g. pick-and-place, the default first task all models can do, worths ~$1 per box at a sorting factory. At home it’s close to 0 unless you pair it with more complex tasks. In a data center, “insertion” task alone can justify six-figure contracts.

Pick your battle wisely, with consideration of capital, usage and roi.

[York Yang @YorkYang5050]

https://t.co/HzlWd2QQ3X

[Paige @paigeinsf]

@KASHIFALIKHAN94 Ya can never turn noti off

[Paige @paigeinsf]

Catching up with a friend who transitioned from Amazon to one of the biggest AI companies lately

Hours be like 60 hours, on call from 7am to sometimes 2am, and weekend standby

“Models update too fast,” she says, “and we are a service company.”

Makes so much sense and many such cases

[Paige @paigeinsf]

@oyhsu I’ve been experimenting with robot financing, especially around settlement gaps across geographies. But I think the more interesting problem is how to spread upfront capex over time and tie repayment to revenue as robots hit deployment milestones.

[Paige @paigeinsf]

AGI and ASI will arrive more like an economic depression than an overnight singularity.

There won’t be a clean before-and-after moment. We’ll only know we crossed the line in hindsight.

Getting closer and closer.

[Paige @paigeinsf]

@wilsonyuan007 Hi sure! let's dm

[Paige @paigeinsf]

https://t.co/4tB4HdywbD

[Paige @paigeinsf]

@odysseus0z it's interesting we think about agents in humanly organizational structure still

[Paige @paigeinsf]

@_simonsmith Speed can be addressed by different harnesses, but what I found interesting is Astra is better at long horizon RL tasks compared to tasks like language annotation

[Paige @paigeinsf]

@Crazyhathello great work!

[Mitchell Hashimoto @mitchellh]

I can’t stress enough how little an idea matters compared to the agency of the people executing the idea.

I have had the privilege of knowing and sometimes even working with some of the most successful people (by various metrics).

The difference between mediocre and excellent work and outcomes is predominantly one of agency.

In practice this means: they dont wait for things to happen to them they go out and make things happen for them.

They don’t wait for someone else to do something, for someone to teach them, for someone to give them the path, etc. They just go out and find a way to do it.

I think the single biggest superpower these people have is the realization/belief that the world around them is completely mutable. Most everything that happens is because a person made it happen.

I used to tell people to look around the room you’re sitting in. Look at everything. Every noun. It almost all exists because a person willed it into existence. Nothing is stopping you from doing the same.

I see people online all the time dismissing someone else’s success because “I had that idea first” or whatever. I mean… yeah? If so then the difference is… you. So a bit of a self own whenever I hear that.

Number one tip: act with agency.

[Paige @paigeinsf]

@LusciousPear Indeed

[Paige @paigeinsf]

@isthatdebbiej Thanks Deborah! What benchmarks you are looking at?

[Deborah Jacob @isthatdebbiej]

For the past month, I’ve been exploring the physical AI eval space. I’m increasingly bullish on deployment evals, and every day I’m more convinced they are the essential to close the demo-to-deployment gap.

Evaluation is hard because the real world changes constantly and there are many factors like hardware, policies, human intervention etc. that affect the outcomes.

The most challenging part is understanding why, under what conditions, and whether that same results can be reproduced in the next deployment.

[Paige @paigeinsf]

@rmn @MunariAI Would love to try it out! Do you have language annotation as well?

[Paige @paigeinsf]

@antopatrex1 Agreed- also commercial level deployment is another level of scrutiny from my experience: https://t.co/AoS5h4InUP

[Paige @paigeinsf]

https://t.co/4tB4HdywbD

[Paige @paigeinsf]

@stephzhan @nvidia Agreed! Wrote about this myself: https://t.co/AoS5h4InUP

[Paige @paigeinsf]

https://t.co/4tB4HdywbD

[Paige @paigeinsf]

imho the biggest hidden benefit of solo + a few agents is near zero communication overhead

Agents follow instructions better than most human counterparts now, and persistent shared context means every new agent "hire" onboards instantly via in-context learning instead of weeks of video calls and file sharing lol

[Paige @paigeinsf]

@adcock_brett Hi Brett, we have 100k users in our community from @FabricFND in similar regions of your map. How can we use your app and contribute data?

[Paige @paigeinsf]

@TongheZhang01 it's great at perception cognition. honestly stunned! but not so much at language annotation