I'm guessing you can still find Coral TPU-based boards somewhere but not sure what the support for these will be now that the focus is shifting. Coral TPU also uses subset of tensorflow and its nice to see that the open standard is targeting jax and torch.
Rodney Brooks is widely recognized and celebrated roboticist, ran CSAIL for a while. iRobot as a company created a new market and managed to put a functional household robot out there, whether Chinese alternatives ate the share of it is largely irrelevant to his argument on humanoids, which I find to be completely reasonable.
There is a wide gap between academic and entrepreneur/business operator though. He was one of three co founders of iRobot the company l, is not even credited at the inventor of it, and iRobot has largely been a one hit wonder. They created a category but the underlying product largely was unimproved until China started dominating.
His other business was a failure and his third current is in a crowded marketplace. Humanoids are the minority in warehouse automation.
Who is even talking about his argument on humanoids? What does that have to do with my comment. My response was on a comment praising his triple success in business and I am questioning that definition of success.
You say one hit wonder, but iRobot's PackBot and other gov-focused efforts were/are very profitable, though the company had to spin off the defense arm.
What are the other efforts? Packbot was sold off for $45mm for the time a company with a market cap of $1.6bn. Again the PackBot could be great but would you call a product that had low $ contracts to the government that eventually was sold for peanuts a wildly successful company? I don’t consider iRobot to be a wildly successful company.
Endeavor Robotics was recently sold for $382MM. So clearly it was actually valuable.
Are you aware that from an investor perspective, creating a household brand and high IRR though IPO is considered a wild success? All companies eventually decline. By your metrics GE wasn't a wildly successful company either. Or AOL. Or Yahoo. Or Myspace. Every one of those companies eventually declined and your return would be poor if you held on to your investment over their entire arc.
Are you aware you are picking the wrong numbers. $382 is post iRobot when endeavor sold to Flir. Why would you even use that number. PackBot sold for $45mm. That’s a decade worth of work and investment for $45mm. I think you might be narrating a bit too much and putting words into my mouth. I don’t know the IRR for iRobot as I was never part of the funding rounds while private or an insider while public. That said it was a stagnant public company that had a jump around Covid time which has come crashing back down. Maybe those folks like yourself, based on the way you speak, invested early in the 90s in iRobot and were part of that IPO.
I am not attacking the man only a perspective that I am not sure investors would see him as a wildly successful entrepreneur.
I might be wrong but isn't the HUD the author suggesting for coding is basically AREPL? For debugging I can see it work, but chatbox and inline q&a I feel has awider application.
On a wider note, I buy the argument for alternative interfaces other than chat, but chat permeates our lives every day, smartphone is full of chat interfaces. HUD might be good for AR glasses though, literal HUD.
Trying to make interpretability research practical. A bit early for the demo, but I am getting some interesting results for large multimodal models in terms of their reasoning.
I don't know, there has been so many overhyped and faked demos in humanoid robotics space over the last couple years, it is difficult to believe what is clearly a demo release for shareholders. Would love to see some demonstration in a less controlled environment.
I suppose the next big milestone is Wozniak's Coffee Test: A robot is to enter a random home and figure out how to make coffee with whatever they have.
Is that a real 5-10 years, a research 5-10 years[0] or 5-10 years of "FSD in the next 6 months"?
The demo space is so sterile and empty I think we're still a loong ways off from the Coffee test happening. One big thing I see is they don't have to rearrange other items they have nice open bins/drawers/shelves/etc to drop the items into. That kind of multistep planning has been a thorn in independent robotics for decades.
Imagine they bring one out to a construction site and they treat the robot as a new rookie guy, go pick up those pipes. That would be an ultimate on the fly test to me.
Picking up a bundle of loose pipes actually seems like a great benchmark for humanoid robots. Especially if they're not in a perfect pile. A full test could be something like grabbing all the pipes, from the floor, and putting them into a truck bed, in some (hopefully) sane fashion
I have my personal multimodal benchmark for physical robots.
You put a keyring with bunch of different keys in front of a robot and then instruct it pick it up and open a lock while you are describing which key is the correct one. Something like "Use the key with black plastic head and you need to put it in teeths facing down"
I have low hopes of this being possibe in the next 20 years. I hope I am still alive to witness if it ever happens.
very good question, now we are mainly focusing on building the foundtion for multimodal perception and atomic action taking. Of course, integrating the trace-of-mark prediction for robotics and human video data enhances the model's medium length reasoning but this is not sufficient for sure. The current Magma model will serve as the basis for our next step, i.e., longer horizong reasoning and planning! We are exactly looking at this part for our next version of Magma!
iPhone has this nice accessibility feature where you can greyscale the screen, this along with putting the phone away in a distance that I would have to get up and walk to it made a huge difference in frequency of usage.
Yes I found the grayscale surprisingly effective. An additional tip is to bind it to the triple-click side button accessibility shortcut, so you can quickly enable/disable it if you need colour for something momentarily.
How do you define and quantify ‘calibration’ in this context? Is it purely based on aligning recommendations with explicit user preferences, or are you also trying to infer latent interests?
