May

11

I've Covered Robots for Years. This One Is Different
From sorting chicken nuggets to screwing in lightbulbs, Eka’s robotic claw feels like we're approaching a ChatGPT moment for the physical world.

Dendi Suhubdy offers deep analysis:

I have put some thought into that:

The Perception–Planning Gap: What's Actually Hard About Visual AI in 2026

The thesis I want to defend in this piece is structural: perception in pixels is largely solved at the representation level; perception for action is not, and the gap between the two is the central unsolved problem of visual AI in 2026. Frontier vision-language models can pass medical-board questions and explain radiographs at attending-physician level. Frontier robots, after a decade of foundation-model progress, still cannot reliably load an arbitrary dishwasher. Moravec’s paradox is not a quaint historical observation; it is the daily lived experience of every embodied-AI lab.

Big Al responds:

Thanks for sharing. Much I don't understand, but when I scroll down to the summary, I think I get the general ideas. I hadn't thought about the challenge of having effective testing/evaluation standards.

Asindu Drileba writes:

I thought this was going to be about humanoid robots (like Figure & Optimus). That still have very many problems. Industrial robotics how ever has always been making mind blowing, but quiet progress.

Good case studies are Amazon (robots operating in a warehouse about 10 years ago, its now way better), and Tesla, also 10 years back.

The reason is that the edge cases of these industrial robots are very few and can be comprehensively thought about. I think this robot ChatGPT moment already occurred in industry & manufacturing.

Feb

18

The old canard, “You never go broke taking a profit” was not coined by a winner, that’s for sure.

Big Al adds:

Markets are an excellent venue for cultivating outcome bias and hindsight bias. It's too easy to look at a chart and think, "Why didn't I buy *here* and sell *there*???" Then you turn to face the future, and the future is blank.

Jeff Watson writes:

People who watch televised poker with the hole cards exposed and % to win, frequently make judgments of the players performances based on the complete information they see vs the incomplete information the player is working with. It’s easy to be an armchair quarterback.

Dendi Suhubdy comments:

True. You could however predict the bluff rate, and percentages from past play, which can be used to win a game. New outlier players are hard to predict because lack of data, you need some sort of one-shot learning there. Hard.

Zubin Al Genubi responds:

The best buys are at the point of maximum pain and uncertainty.

Archives

Resources & Links

Search