The profit from a scheme like this would likely be in the high tens of millions of dollars.
The poker game itself in high-roller situations could be a million plus per night depending on the stakes.
Then there's the whole "you owe the Mafia" angle with NBA players and coaches. It's a pretty clear line to the Mafia making tens of millions of dollars on rigged NBA games.
First, people who are banned from online bookies use "horses" or other not-banned players to place bets for them.
Second, the FBI is targeting real world Mafia members, who will typically be the bookies taking action from others. If they know in advance, through blackmail or collusion that an NBA player or coach will throw a game, they can exploit this versus their entire betting pool for massive wins against the suckers placing bets with them.
Your millage may vary, but I just got Cursor (using Claude 4 Sonnet) to one shot a sequence of bash scripts that cleanup stale AWS resources. I pasted the Jira ticket description that I wrote, with a few examples and the script works perfectly. Saved me a few hours of bash writing and debugging because I can read bash, but not write it well.
It seems that the smaller the task and the more tightly defined the input and output, the better the LLMs are at one-shotting.
I’ve had similar experiences where AI saved me a ton of time when I knew what I wanted and understood the language or library well enough to review but poorly enough that I’d gave been slow writing it because I’d have spent a lot of time looking things up.
I’ve also had experiences where I started out well but the AI got confused, hallucinated, or otherwise got stuck. At least for me those cases have turned pathological because it always _feels_ like just one or two more tweaks to the prompt, a little cleanup, and you’ll be done, but you can end up far down that path before you realize that you need to step back and either write the thing yourself or, at the very least, be methodical enough with the AI that you can get it to help you debug the issue.
The latter case happens maybe 20% of the time for me, but the cost is high enough that it erases most of the time savings I’ve seen in the happy path scenario.
It’s theoretically easy to avoid by just being more thoughtful and active as a reviewer, but that reduces the efficiency gain in the happy path. More importantly, I think it’s hard to do for the same reason partially self driving cars are dangerous: humans are bad at paying attention well in “mostly safe and boring, occasionally disastrous” type settings.
My guess is that in the end we’ll see less of the problematic cases. In part because AI improves, and in part because we’ll develop better intuition for when we’ve stepped onto the unproductive path. I think a lot of it too will also be that we adopt ways of working that minimize the pathological “lost all day to weird LLM issues” problems by trying to keep humans in the loop more deeply engaged. That will necessarily also reduce the maximum size of the wins we get, but we’ll come away with a net positive gain in productivity.
Same. I interface with a team who refuses to conduct business in anything other than Excel, and because of dated corporate mindshare, their management sees them more as wizards instead of the odd ones out.
"They're on top of it! They always email me the new file when they make changes and approve my access requests quickly."
There are limits to my stubbornness, and my first use of LLMs for coding assistance was to ask for help figuring out how to Excel, after a mere three decades of avoidance.
After engaging and learning more about their challenges, it turned out one of their "data feeds" was actually them manually copy/pasting into a web form with a broken batch import that they'd give up on submitting project requests for, which I quietly fixed so they got to retain their turnaround while they planned some other changes.
Ultimately nothing grand, but I would never have bothered if I'd had to wade through the usual sort of learning resources available or ask another person. Being able to transfer and translate higher level literacy, though, is right up my alley.
Feetech is selling actuators which are mechanically R/C type servos, but have a bidirectional computer interface allowing the control computer to find out what's happening at the servo.[1] This isn't new; Dynamixel has been doing it for over a decade. But not at this price point. This Feetech servo is $17, while Dynamixel units start around $70 and go much higher.[2]
The parts list has "need to be strong" for many of the small parts, but they are 3D printed PLA plastic. That's the low end of 3D printing. None of the videos show the hand handling anything.
So this is really the proof of concept model. If there's enough interest, someone could make the parts by injection-molding of something better, such as polycarbonate or glass-filled nylon. The total plastic volume here is so tiny that the plastic cost is negligible, and there's no reason not to use a high-quality engineering plastic.
Nobody seems to do hobbyist injection molding much. TechShop had a desktop injection molding machine, the CNC milling machines to make molds, and even Autodesk Moldflow to design them. But nobody used those tools. A few university maker spaces have similar machines. Because most of the world's plastic stuff is made by injection molding.
Mold design is still difficult when the parts aren't dead simple. The software I've seen is okay with the simple stuff, but once you get even a little more complex you have to understand simultaneously how to design good parts for molding and how to design good molds, both of which are heavily dependent on the type of plastic you're using and the size of the press you have access to. Not to mention how to machine good molds from metal, which is a challenge all on its own due to surface finish and tolerance requirements (and weird geometry that makes the CAM choke...)
In other words, we're not really there yet to bring that activity into the hobby realm. But I hope that we're not too far away.
They're little linkage parts, mostly flat.[1] Some of those holes are bearings. None of those parts are hard to make, but they need to be strong. They could be made by CNC machining, or in quantity by injection molding, or stamping. But tiny working parts in 3D printed PLA will be too flimsy for that hand to do much work.
Totally fixable problem. Then this hand can go to work.
If this thing catches on, someone might sell an upgrade kit with stronger parts.
The designer is already considering a servo upgrade.
I haven't looked too carefully, but I suspect those "needs to be strong" parts could mostly be cnc machined with an inexpensive hobby CNC like a 3020, from aluminiumn sheet/bar stock.
That's what I was thinking, too. All but one of those parts can be easily milled from flat plate. The yoke end is a bit harder, but yoke ends are cheap standard parts and there's probably an off the shelf solution.
The ball joints they use are standard R/C car parts. Those are available in in a stronger form, nylon with a brass ball, for EUR 0.53.[1] It won't be hard to strengthen this design so it can do useful work.
Smiles from Detroit. I would look to press those parts, since most contact surfaces are potentially convex and variant-depth parts appear superfluous to requirements. With a shared punch and die you could pull the nest off a laser, press them all in parallel, and break them after the parallel press-forming. Should be cheaper than injection because the cycles can be faster, the material can be stronger, there's only one die, and the unload will be easier?
Yes. Here are very similar parts produced with a punch press.[1] In Silicon Valley. Or in China.[2]
This sort of thing is probably about $40,000 for the first part, $0.05 for each additional part. Designing and making the custom dies is expensive. Banging out the parts is cheap. Mass production works.
People have been making end effectors using hobby servos for ages. These servomotors are designed for use in an RC aircraft, they're light, cheap, and expendable.
Industrial needs care not about weight, care less about cost, and care a great deal about capability, repeatability, and reliability.
This is a cool project for a hobbyist but it's not meant to be a serious industrial machine.
Edit: what is with this thread? Lots of very generic positive comments here but not much thinking about what this is actually useful for.
What you're missing is: today, we're nearing the point where actual general purpose robots become viable.
Which means: the purpose of a robot is no longer to sit at a factory line and precisely execute the same exact motions on repeat 24/7. The purpose of the next generation of robots is to learn generalized behaviors, adapt to circumstances, and carry out circumstance-specific actions with active sensor feedback. Which means completely different requirements for effectors.
Which means: repeatability can go get fucked, for one.
Humanoid robotics research was pretty popular in the early 2000s already, with remarkable, reproducible results not only in videos. It’s definitively more present in the media now.
It’s one of best designs I have seen, I admit. But for that price you cannot get absolute encoders outside the motor, reliable force/torque sensors (think picking up a strawberry), tendons (thread below). It might be too limited for research and real-world projects unfortunately.
Depends entirely on the application I guess, but you'd get further for sure. Either way you are now confronted with more complex cable management and maintenance, and additional weight. You'd at least get around gravity compensation with that approach. Personally, I think you need a lot more sensors and complex sensor fusion to achieve acceptable dexterity to be useful outside of an industrial context and this comes at high costs
Poker is interesting. I think these videos do work in our current course generation process. However, I do think some subjects like poker need custom tooling around the course to really make the learning experience great. For example, access to solvers or actually playing a hand on a table is a part of the course experience as well. Chess is another one that falls in this special bucket imo. Some of this tooling is on the roadmap!
Yes! I wish there were better tools for getting sound out of the terminal.
I got bored waiting for long running scripts to finish one day and wrote this bash one liner.
Now I can run a command and put a pipe and the zzz alias on the end and it will tell me when it’s finished.
> alias zzz='echo "zuhg zuhg\n" "work work\n" "lowk taar\n" "jobs done\n" "sawobu\n" "dabu\n" "i can do that\n" "be happy to\n" "OK\n" "no time for play\n" "my life for the horde\n" "what you want?\n" "master?\n" "dabu\n" "for the horde\n" "okie dokie\n" "hragh\n" "something need doing?\n" "ready to work" | shuf -n 1 | say -vRocko'
The poker game itself in high-roller situations could be a million plus per night depending on the stakes.
Then there's the whole "you owe the Mafia" angle with NBA players and coaches. It's a pretty clear line to the Mafia making tens of millions of dollars on rigged NBA games.