与「robot」相关的搜索结果

OK let’s try this: ———————- Pretend you’re 5 and the algorithm is a smart robot that picks which toys (posts) to show you in your toy box (For You feed). The robot doesn’t just count how many times people already played with a toy (likes, reports, etc.). Instead, it looks at you and guesses: • “How likely is this kid to like this toy?” • “How likely is this kid to report this toy as yucky?” • “How long might this kid play with it?” Those guesses are little numbers between 0 and 1 (probabilities). The “weights” are like special multipliers the robot puts on those guesses: • A like-guess might get multiplied by a small number (like 0.5). • A report-guess might get multiplied by a big negative number (like –234). Then the robot adds all those multiplied guesses together to give the toy a final “yay or nay” score and sorts the toys by that score. Why the numbers look so big and why people get confused Reports almost never happen, so the robot’s “will you report this?” guess is usually a tiny number (like 0.0001). To make that tiny guess still matter a little bit in the final score, they give it a big multiplier. People sometimes look at the big number and think “1 report = 468 likes!” 
That’s wrong. It’s not counting real reports or real likes. It’s only multiplying the guesses the robot made about you. If the robot thinks you’re the kind of person who almost never reports anything, even a big weight on “report” barely moves the score. The whole thing is personalized to how you usually behave. The people who wrote the code even put big comments in the GitHub repo so no one (and no other robots) gets confused about this again. That’s the whole idea in kid words: the weights turn the robot’s personal guesses about you into a ranking, not a simple “count the hearts and flags.”

How much would someone pay for a partner who never leaves? The humanoid in this clip is presented like a luxury object, but the promise goes far beyond appearance. The buyer is really being offered attention, conversation, and the feeling of having someone nearby. That is where robot companions become uncomfortable. A car or watch does not pretend to care about you. A social robot is valuable only if the interaction feels personal. Would knowing it was programmed make the connection meaningless to you?

The future of Optimus will be decided by the egg, not the squat. Tesla's Gen 2 demo cuts from a humanoid moving a fragile egg to squats and finger gestures. The squat looks impressive. The egg is the hard part. A useful humanoid will have to handle soft, slippery, unpredictable objects thousands of times without a reset or a human stepping in. The next important Optimus video is not another trick. It is one ordinary task repeated for an entire shift. What would convince you first: an eight-hour warehouse test or a robot doing chores in a real home?

Just took a Tesla Robotaxi ride in an area of Austin where the normal wait in the past was up to 20 minutes and my wait was only two minutes and this was consistent every time I did this!

A robot cheetah now runs 45 km/h, faster than Usain Bolt's top speed. It's not the strangest one. A robot spider skeleton comes off a 3D printer. 26 servo motors. $1,385 with skin, $250 without. A robot manta ray took 2 years and 40 fin designs before Singapore researchers landed on flexible PVC pectoral fins. It swims at twice its body length per second. 10 hours underwater before it needs to surface. A robot dragonfly from Germany weighs 175 grams. 27-inch wingspan. Flies forward, backward, sideways, hovers, just like the real thing. Boston Dynamics built a dog that carries 500 kg and hits 25 km/h. It can open doors on its own. A 6-legged crab robot drops 200 meters underwater on 30 joints and grips objects with its front claws. A robotic kangaroo weighs 7 kg and jumps 80 cm, recycling the energy from each landing into the next leap. An AI puppy remembers what earns praise and repeats it. Eyes are screens. $2,899. None of these were built to look pretty. They were built because evolution already solved the problem first.

7/ 🧭 Dyna-2 in one line It's not "robots should watch videos." It's this: physical intelligence has its own scaling variable, and human experience is the most scalable one we've got. 1M hours is just the start. Full technical report 👉

6/ 🌏 "Robots can only learn from robot data"—busted A thousand robot demos take real effort. Learning from millions of hours of people cooking, folding laundry, assembling things is a different order of scale. The future's biggest robot dataset may not come from robot experts.

5/ 🔧 A few hours of data → real hardware Same human pretraining; robot post-training takes just hours of data—scores climb 20%→53% as pretraining scales. 10 min of teleop fine-tunes it to unscrew a cap with two dexterous hands. Zero-shot at a new customer site: 46% → 87%.

1/ 🧭 Robots just found their scaling variable Dyna Robotics dropped Dyna-2: a world-action model pretrained on 1M+ hours of human video—about 170 years of nonstop waking experience. The headline isn't the robot. It's the scaling law behind it—still climbing, no ceiling yet 👇