Ask HN: I quit my job over weaponized robots to start my own venture
ZH version is available. Content is displayed in original English for accuracy.
I’ve decided this is the right time to go back to entrepreneurship. We're at an incredible moment for embodied intelligence, but I feel the tools and workflows we use to interact, monitor, and control these platforms are still lagging behind.
I'm currently exploring a couple of projects around how we build, test, and interact with robots. As part of my customer discovery phase, I'm trying to gather raw data on how roboticists and developers actually work day to day and what their main pain points are regarding control interfaces.
I put together a very short survey (3 mins) to validate some ideas. If you work in robotics, embedded systems, or just tinker with hardware, your input would be incredibly valuable:
Survey link: https://forms.gle/3Nm76wkeT5CMt23c8
I'm also open to discussing the ethical lines in modern robotics or anything related to ROS2 / HRI in the thread. Thanks for reading!
Discussion (71 Comments)Read Original on HackerNews
I'm on a similar journey. I took flight 6 weeks ago and built a turn based board engine for human/agent delivery teams called Keel https://www.spoke.sh/keel. The grand vision is to apply the board engine as a control mechanism for work to be done and verified in deployed robot fleets.
I am forced to accept the popularity of ROS but I find it to generally be a terrible experience. Are you considering an alternative? Have you used foxglove?
I am definitely looking into Foxglove! It seems to solve many of the transport/protocol headaches, but I feel like there's still a massive gap in how we actually interact with the robots day to day, especially when you are not glued to a desktop monitor.
I'd love to hear more about your experience. What specific part of the tooling drove you crazy enough to start building an alternative?
(Also, if you are open to a quick 15-min chat to share "war" stories, let me know!)
Helping people enhance is a good thing!
Very strong disagree; a lot of people is objecting. A job on an assembly line may be "bad" for somebody, but for somebody else can be a lifeline, if they won't be able to find another job soon enough and/or in reasonable conditions. Long-term, the job market can rebalance (and if unemployed people are supported in their education, it's great), but short-term displacement is a serious issue.
I'm glad I did it though. We have to few years on this earth to spend our energies hurting others.
Don't you live in a nation state that uses violence to maintain the order that you've come to enjoy? Here's a harsh dose of reality for ya, suffering is unavoidable... the trick is convincing the worker class that it's easier to just cooperate
therefore it makes no sense to consider one's own role in producing, mitigating, or directing suffering in the world
i am very smart
/s
I hope you are able to convince some of your colleagues to do likewise.
With robotics and AI, it feels like there are a lot of directions it could go that would lead to higher quality of life and not just temporary advantages for killing other people.
Won't be surprised to see hundreds of thousands of humanoid robots strapped up with explosives running to their target or some of them flying to their target with drones attached.
https://www.youtube.com/watch?v=O-2tpwW0kmU
https://www.businessinsider.com/ukraine-exoskeleton-test-bat...
Cakes exist and I even like them, and I do not choose to work at a bakery.
All the drone warfare developments remind me of the introduction of tanks during the first world war and perfected by the second world war. In the space of a few years they changed warfare. Then planes changed warfare again. Now drones. Makes you wonder what the next thing will be
Which of course leads to point 2: it's very easy to take a moral stance on weapons when you don't think you're in any danger, nor going to be doing any of the fighting otherwise.
Rather than blowing up a school full of little girls, you could deploy a swarm of thousands of fast-moving cat-sized robots armed with tasers and bolas to identify and capture targeted enemy leaders.
I've been using AI heavily to do this, so everything is in ROS2 since it's "standard" and AIs have pretty good training for it. I can see how it's annoying and suboptimal if you're writing manually and after a more integrated system, but it's been pretty good for getting up and running because it's "standard" and kinda plug and play. I see why you'd want to rewrite it for production, the endless processes and nodes and startup processes can get annoying
One of the more useful things I've done so far is actually not robotics related directly, it's a Godot based "game" with a ROS bridge that lets me drive the robot from Foxglove, which I will eventlly get a vlm based agent to drive. Seems much easier and faster than Issac Sim for getting started with.
Using a survey like this is IMO not ideal though.
One doesn't need to compromise on one's values to earn.
I was in the past in the position of working for a corporation I personally consider to be vile, damaging to the world and society. Took me about 3 years to move elsewhere. I was not in the position to just quit, both due to finances and due to visa requirements.
I don't fault the common man for having to put up with things. But I will commend those that have the fortitude to at least turn and walk away.
There are many open source solutions out there: https://alternativeto.net/software/google-forms/?license=ope... I recommend if you can choose any of privacy friendly options, thanks and have a nice day.
I’d be wary of a founder with such bad NIH
It would be like writing your own email servers or calendar software. It would be a distraction at best.
Write a form in .md (even tell an llm to do it) and just put it online.
Web forms are simple like that slack notification thing is simple.
Unfortunately it doesn't matter, some else will go ... just look at the ukr war.
I knew about BD's history with DARPA, of course. The issue was that my company was doing some actually really interesting non-defense work, and then decided to pivot and mount teleoperated weapons on these platforms for a new demo. That’s when I submitted my resignation :)
Human emotions and reasoning could be internally inconsistent and conflicting, yet everything is as it should be, counter-intuitively.
But as expected, others have taken their place [2]. Guilt-tripping a single non-monopoly proving useless again.
[1]: https://bostondynamics.com/news/general-purpose-robots-shoul...
[2]: https://www.digitimes.com/news/a20260323PD219/military-bosto...
That’s just marketing bs in the same vein as “Your privacy is very important to us.”. It means nothing.
The "no more military weapons" statement seems to have been after they were acquired by Hyundai.
Boston Dynamics' business is, basically, "mobility platforms." After all these years the basic development is all done; now they're pivoting to commercial markets.
There's no real difference between a "murderbot" and, say, a police riot-control platform, a fire-fighting platform, a forestry platform, etc.
They might not be explicitly developing weapon packages any more, but there are plenty of other companies who will be happy to take the money to build them onto Boston Dynamics' platforms.
He said he worked with their hardware, not that he worked for Boston Dynamics.
Entirely possible to be working with a platform provided by Boston Dynamics at a company that is not engaged in weapons development.
Sure, Boston Dynamics is a bit more obvious there, but merely having DARPA funding doesn't mean it's about killing people.