Post

@Rewkang Very good one

@Rewkang ⤴️ Right before everything moves

@Rewkang robotics thesis summaries are the new white papers next stop: a robot that writes its own summary

@Rewkang Thanks for sharing， I've read it carefully and am truly optimistic about the robotics sector it has enormous potential and represents the future trend， We should seek opportunities within it🫡

@Rewkang By the way, can I make a request? Could you update your tweets more often? I've turned on notifications😊

@Rewkang My internal plan is as follows ⬇️ 👇

@Rewkang Solid.
AI finally gives robots a brain.
Huge TAM ($60T labor), cheap hourly cost, accelerating fast.

@Rewkang robotics thesis recaps only help if they stay ahead of the build cycle acceleration just means last quarter points already look late

@Rewkang 🔥

@Rewkang Agreed

@Rewkang *mandatory to understand. This piece really opened some doors that were hidden.

@Rewkang Unitree going from zero to beating human records in 3 months is exactly why this thesis is so compelling. We actually covered it here: x.com/AGTPinsights/status/2089…

@Rewkang Looking Good.

@Rewkang Which areas in the robotics thesis @RoboStrategy is least exposed to (missing exposure)? What about the Chinese robotics industry?

@Rewkang this is a great summary here and I've been following you guys for a while, and wondering what you think about the robotic data space, whats ur hot take here🤔

@Rewkang This may actually get the normies to be pro ai just as much as curing cancer

@Rewkang 1 billion American robots

@Rewkang Due to societal implications of mass employment loss I suspect governments will license commercial robots, limiting the market somewhat.

@Rewkang Timing risk. The main problem is the Chinese hourly wage is $2 / hour. $1.66 is barely a discount and you’re not including integration costs, mistakes, downtime, etc. It’s not clear robotics labor is a slam dunk yet, especially given the current pricing - like quantum.

@Rewkang Agreed

I get the business thesis. What is not clear is the technical thesis: How large the gap is from now until the drag-and-drop dream? A few guesses here to make the point: If Amazon and Walmart have 1 million labor pool, maybe 10% (?) of that is addressable for robots by 2030 (35?) because the jobs they are doing are very repetitive and simple enough for a robot. But the jobs in other factories are not the same warehouse jobs as in Amazon and Walmart. Also, only the labor pool in places where the per capita GDP is $30K is addressable. Labor in many parts of China and Vietnam is still cheap. Then labor in Bangladesh, India and Africa is extremely cheap. Even with those exclusions, the TAM is quite large, so it is still an attractive market. But there is a timeline aspect to this that we should be real about. Then again, if someone had told me in 2018 that there would be these chatting bots that will be as smart as a PhD, I would have laughed at him 😂

@Rewkang The next robotics advantage may come from recovery, maintenance, and reliable deployment rather than demonstrations alone.

@Rewkang This reads like crack if you’re retarded

@Rewkang Two best people in one house, love it

@Rewkang x.com/joshchapmn/status/208963…

@Rewkang What's your thoughts on the data bottleneck for the robotics training?

@Rewkang Once machines possess a brain capable of truly understanding the real world, combined with an ultra-low labor cost of approximately $1.66 per hour, the entire $60 trillion global labor market instantly transforms into a potential TAM

@Rewkang x.com/gerardsans/status/208941…

@Rewkang Completely ignoring the race to the bottom nature of the business

@Rewkang what about data for robots?

The key question is not whether robots are cheaper than a human wage today, but whether they can deliver a lower total cost of ownership at scale. Labour cost is only one variable. Integration, supervision, maintenance, downtime, error rates and utilisation may determine the real economics. That is why deployment data matters more than headline robot pricing. As reliability improves and systems become easier to integrate, the cost curve could change quickly. The winners may be those solving the full workflow, not simply building the hardware 🤖📉 Please read my latest X: x.com/tekkaus/status/208974397…

@Rewkang Why would WMT and AMZN replaces 2M labourers with 2M robots when the labor works 40hrs/wk and robot can do 140hr. The lease case works out to 7$/hr. 50,000$ / (140hr/wk x 52wk)…

@Rewkang the $1.66 hr number is wild but ignores how quickly humans adapt

@Rewkang there is no doubt humanoids will be a massive driver of economic growth. there are also many areas where other robots will outperform humanoids, for example in large-scale infrastructure construction x.com/thelcjones/status/209012…

@Rewkang The chart itself undermines the thesis-if that low point is truly pointless, why are you investing before the spike?