Chat gpt: Task~ Order all cooking oil from Wally World

---

Don't forget about RedWire. Red Cat's Black Widow drone that won SRR, will also integrate with Redwire their LRR won drone. Red Cat is proving to be a powerful integrator. And I am wondering how their partner Palladyne AI will fit into this, being the only US company making actual swarm technology possible where an operator can operate a fleet of drones that all have their seperate task yet communicate with each other. Do with it what you will.

---

Oh I get it now. That if he was able to bring in the Middle East then then the bet the task of ending the government shutdown is nothing to him.

---

So i was an ESL kid in school and lemme tell you "it’s just Language models which often generate errors or incorrect answers." This is a way bigger technical issue with the squishes (especially the only English speaking ones) than it is with the transistors EDIT: Also not to be educating the internet since it is futile, and every time I try some AI snitches me to the feds... if you think the primary use case is entertaining idiots with strings of human legible text a-la [1964 Eliza](https://en.wikipedia.org/wiki/ELIZA), you will obviously conclude it is a bubble due to what is called a ceiling effect, morons will think it's peak tech at all times. Look at Quantum-Classical hybrid computing where AI based physics engines act as ECC for qbit errors, or AlphaFold2 (nobel prize winning prediction for protein shapes that gets you a MRNA vaccine formulation in a few weeks), FFS - the rockets are landing themselves vertically - the signal and systems/control theory and sensor fusion required to perform that task are not something you are gonna work out on pen an paper and hard code. I know people aren't paying attention, but electrical rocket porpulsion is a thing, and the only way you get that going sustainably is by putting nuclear reactors in the rockets... no way I'd be comfortable with fissile material on a rocket unless it was AI controlled, people push all the wrong buttons. https://preview.redd.it/e3xmj3q7kpuf1.png?width=400&format=png&auto=webp&s=d66e12b208e3cf76f70e2230b91da650db671241 Them google people are about to create digital twin simulations for biological systems [https://www.youtube.com/watch?v=-HzgcbRXUK8](https://www.youtube.com/watch?v=-HzgcbRXUK8) Again, not a bubble in a classical financial sense, more like the money is useless if you have a calculator that can solve more problems than money can.

---

Why is task in French

---

Because there was no other company doing anything close to this, and asts had been at it for 7 years already. After a couple of hours of DD I knew it was a winner, and that they were far ahead of anyone else. It was literally called an impossible task to get broadband to your unmodified phone, and asts cracked the code. Att signing on was a huge signal that the tech worked like asts said it does. I bought my first 500 shares before Att signed on, they went up 250% and I was hooked. Bought every dip and dcaed every paycheck for a year and a half. Trimmed or flat out sold other positions to fund my asts position. Still doing that because it’s the only real 3-10x I see as a possibility. Also things like the kook report, weekly updates from extremely smart people with their entire life savings in asts, and a ton of other extremely thorough dd. When I got in there were very few articles about asts outside of these reports, so I was able to read every single article and report that was on the internet. Now so many articles come out every day you can’t read them all. Spacex is their closest competitor, and even with unlimited resources they are a minimum of 2 years behind, which is enough time for asts to have a global monopoly for a year (and better tech on top of that). The more I read into it the more comfortable I felt putting more money in, until I had 1/3 of my entire net worth in. This took me more than a year and a half, im up to 5000 shares now. If I didn’t have a family I would have gone all in. I expected these kinds of returns in the next 2+ years, not this quickly. But I did see the pltr type run up as a possibility around 6 months ago and started piling on at an accelerated rate (much to the discomfort of my wife at times hahah)

---

I think, no matter how good a humanoid general purpose robot gets, a special purpose non humanoid robot will always be more efficient for the task. People keep trying to make humanoid robots work because of sci fi and fuckbots but it just doesn't make sense. An automated warehouse doesn't need bipedal spergs in it. If a company *can* make humanoid robots work in some capacity that's good PR but they probably still should and will make other things. Except for Tesla because Musk just wants to do sci fi.

---

Apparently its 200 troops - a task force specializing in logistics 🤷🏾‍♂️

---

I don't really disagree with you. In the short term, automating equipment built for specific tasks (cars, floor sweepers, forklifts, mining equipment, etc) is going to be more productive and profitable. but the promise of humanoid robots is based on the idea that our entire world is designed around humans. Once you're able to build a humanoid robot that is dextrous and smart enough to be able to do basic manual tasks, then it immediately becomes incredibly versatile. Humanoid robots can interact with the machines we already have. Take a fast food kitchen for instance. There are a bunch of machines designed for a specific task (freezer, mixer, fryer, dishwasher, register etc) and most of what fast food workers do is move things from one machine to another over and over. Once you build humanoid robots that can do basic tasks like open a freezer, remove food, close freezer, put food in fryer basket, etc. that is a huge amount of man hours in our economy. It's going to be a while before that replaces workers, but eventually it's going to happen

---

Having used Boston Dynamics products, I really don’t think they’re all that great. They’re fine if you need remote surveillance or something, but expensive compared to drones. They aren’t very useful at doing things on their own. Developing software to have them perform a task is expensive and limited to simple repetitive tasks, and in the end it’s difficult to make the financial case close. The software is just not there. For humanoid robots to catch on they need to be capable of being easily adapted to a wide range of tasks and they probably need to come down a good bit in price too.

---