@niccruzpatane Let’s be real here: robotaxi day was in June of 2025.
They said they’d cover half the US by end of year.
We’re now more than a year after the original announcement, it covers almost nothing.
Was at the BJ robotics games, note that many teams participate but what gets shared here is the top 0.01% of all content.
The vast majority are clanky as hell, essentially none seem to be autonomous in any way.
people are arguing both sides: that the chinese have solved dexterous manipulation (not sure if true) and that americans have robots that can run fast (not sure if true). my point still stands about which promotional material surfaces from each country
None of the tens of thousands of Chinese robots have any autonomous capacity at all.
US robots build it from the ground up.
I’d know, I was at the robotics games in Beijing today, talked with the teams working on them.
I don’t think Americans understand how behind we are in robotics.
- Most “American” humanoids e.g. Figure/Optimus using Chinese parts.
- Not shipping a single bipedal humanoid.
- US is producing small batches of 10-20 for R&D while China is shipping tens of thousands.
@levelsio This is an oversimplification.
Technical analysis has weight in investment decision. Maybe 10%. Probably less
Therefore it’s useless for anyone but the people who spend their lives in stock trading, hoping to gain a few % advantage.
In case you missed it…
A fully expendable V3 Booster, Starship can still make Mars arrival with 40–50 t of propellant left.
Aerocapture does the heavy lifting (almost zero prop cost), then you get a full week in orbit to image landing sites, watch for dust storms, and pick the perfect moment.
After that, deorbit + landing still leaves a 5–10 t margin.
One-way test flight. No tanker required.
Expend it. Send it. 2026.
@Sonderling47@SpaceX@StarbaseTX Do we need or want a SpaceX Space Shuttle? You could most likely develop one inexpensively. Especially since one was already invented and used successfully. It was just very expesive. My guess is if SpaceX developed a shuttle it would cost as much as a commercial jet!
Added the SENTINEL-3 satellite as a new source instrument for more frequent heatmap updates over intense wildfires.
(thank you @esa !)
(also thank you @NASA, you're also giving me loads of data!)
Built IGNIS, a dashboard that combines satellite data from around the world to track wildfires in Europe.
https://t.co/paWpREsVyx
Used Grok 4.6 Fast exclusively, first proof of concept in less than thirty minutes, full website in less than 6 hours.
Very impressed by its performance & speed.
Bad take :
- how many human videos in general “move you” or “linger in your memory”?, in %, for a given time period.
- how many AI generated videos do you watch in that same time period?
Not to say AI videos are as good, but some are. Higgsfield’s AI movie is genuinely okay. It really has its moments.
Also, video models have become decent only a few months ago.
rapid progress, yet I cannot name a single AI-generated video that has moved me and lingered in my memory as something honestly worth watching, nor even a single good adaptation of an existing work
at some point "rapid progress" in tech demos is insufficient. will there be art?
Built IGNIS, a dashboard that combines satellite data from around the world to track wildfires in Europe.
https://t.co/paWpREsVyx
Used Grok 4.6 Fast exclusively, first proof of concept in less than thirty minutes, full website in less than 6 hours.
Very impressed by its performance & speed.