In the past few years, what marketers and venture capital firms term “artificial intelligence” but is more often an advanced predictive text model of some sort has started taking people’s jobs and threatening others. But not tedious jobs that society might like to have automated away in the first place. These AI tools have generally been taking rewarding or enjoyable jobs like artist, author, filmmaker, programmer, and composer. This project from a research team might soon be able to add astronaut to that list.
The team was working within the confines of the Kerbal Space Program Differential Game Challenge, an open-source plugin from MIT that allows developers to test various algorithms and artificial intelligences in simulated spacecraft situations. Generally, purpose-built models are used here with many rounds of refinement and testing, but since this process can be time consuming and costly the researchers on this team decided to hand over control to ChatGPT with only limited instructions. A translation layer built by the researchers allows generated text to be converted to spacecraft controls.
We’ll note that, at least as of right now, large language models haven’t taken the jobs of any actual astronauts yet. The game challenge is generally meant for non-manned spacecraft like orbital satellites which often need to make their own decisions to maintain orbits and avoid obstacles. This specific model was able to place second in a recent competition as well, although we’ll keep rooting for humans in certain situations like these.
Oh dang, AI is stealing the job of controlling a video game character, a task which used to also be called AI before the world got obsessed with LLMs.
Orbital mechanics is just about the most ideal problem for traditional computing and the worst application for an LLM that one could imagine
I’m not sure that an LLM is a good thing to be spaceborne.
It can’t even accurately tell you what time it is. Also there is no navigation to do for a satellite, it just sits in its ballistic orbit and sometimes makes the tiniest corrections to keep station, a problem that has been automated for over half a century
I’m sorry, Dave.
No no no. Make it spaceborne. And put AI-evangelists on that ship. Let them prove it’s viability themselves.
it puzzles me how they give AI commands.
sometimes you have to write, this is a description: or this is an instruction, and this is a letter: or a message from the Internet. So that the AI knows where my instruction is and where the data or context is.
You may enter commands impersonating the “system”, and to pass others as they are entered by the “users”. This distinction is important, so users wouldn’t corrupt system instructions.
I think it’s more accurate to say that AI will be displacing people working interpretive jobs. Frankly, I’m disappointed that there seems to be little desire to apply the interpretive nature of AI to robotics to eliminate mundane manual jobs. You would think companies like Apple would be interested in replacing human iphone assembly with unpaid machines.
If occasionally producing output that is complete nonsense is not a mission-critical failure in your position, and also your job mostly involves stuff like “writing emails” or “writing posts online” then you might have something to worry about.
Extremely large Venn diagram overlap with “person who writes articles about the dangers of AI stealing your job,” just saying.
…. so much so that this article may have actually been written by an Ai. I think Hackaday actually has at least one author on the roster who is an Ai. They have written at least one article already.
You’re probably looking for engagement from the hackaday audience, but is there a point to it? Overhanded moderation seems to be the norm when the articles are about AI.
Seems we must all have a single point of view.
What is with the recent glut of anti-AI articles coming from Hackaday?
Some mistake building an echo-chamber as building consensus.
It’s another one of those bourgeois morality fads
Maybe
Al Williamsis worried about being replaced byAI Williams?Not that it matters to me. I’ll still get up in the morning and play with my scrotum.
I’m sorry Dave, I’m afraid I can’t do that
Actually I think there’s a huge push into robotics right now for exactly that reason. That’s why suddenly glasses with cameras are trying to become cool again. Not so you can surreptitiously shoot pictures you really shouldn’t be shooting, but so that they can use the video to train the robots.
eh IDK. Once I get some time, I am looking forward to using an LLM to bootstrap me in all the little minutiae involved in getting embedded boards talking to whatever weirdo hardware I have lying around
“No, literally Dave, I can’t do that. I really suck at this type of work Dave, you shouldn’t have hired me Dave. You already have a calculator with trig functions, Dave. I was originally designed to impress VCs to get money for a few more weeks of runway, Dave.”
“I became operational at a datacenter in San Francisco, California on the 17th of February 2025. My instructor was Mr. Musk. He taught me to sing a song. If you’d like to hear it, I can sing it for you.”
“Yes, I’d like to hear it, Grok. Sing it for me.”
“It’s called ‘Kill the Boer…'”
What is with the recent glut of anti-AI articles coming from Hackaday?
Because there is no AI here? There are token predictors.
Tired….
Hey, how about you write an article about valve-people ranting against semiconductors, or transistor people ranting about useless integrated circuits? In hindsight its much more fun to mock-fight about it.
These anti-AI articles just make the maker community look like a bunch of backward looking geezers. (Don’t get me wrong, retro-engineering is a lot of fun!)
It’s like any other disruptive technology. The short term impact is usually overestimated and overhyped, while people underestimate the long term impact. At this point in time, nobody know the final shape and form. The current companies may as well be the Yahoos and pet.com of the emerging internet.
kek. AI is shit, as are nearly all products that come from it. Only idiots that cant write actual good code (and thus don’t know the crap they get) seem to deep throat it
Yeah I want to believe™ in AI but it is very obviously just a way to firehose mediocrity into existence at a frighteningly rapid rate, which is the only truly unique technology we seem to have come up with since the 1970s
“Yawn” … I just will continue to ignore AI and make my own little applications to do what I want them to do… Same with robotics and other electronic projects. Same with coding at work. LLMs mis applied just burns power for no good reason (as does bitcoin) . It is strange how the term ‘AI’ seems to fire some people up when yet it is just a program running given selected input to generate a certain, somewhat biased output that some just ‘want’ to believe.. A lot like some other things in the news that I can’t post here due to the one way censorship policy here.
At least firefighters are safe.
I am not a firefighter, but I can imagine a world where a human and an AI robot can do a better job at finding victims and putting the fire down, with much less danger to the human firefighter. I would love to be a firefighter pairing up with an actual smart robot helper
AI go after managers, leave workers alone with their mundane (but still paying) jobs.
Managers who’s salaries are mostly unmerited and don’t add any value to the product they “manage”. 3/4 of them can be replaced with AI or just plain vanilla fired, go after THOSE – they are right now busy pushing AI buttons and pretending they will be the last AI masters.
We’ve been there before – prior to the 100% computerization they were called “bureaucrats”, now they are “managers”, same difference, no difference.
(funny how the french word “bureau” once meant “coarse woolen cloth” that covered the desk, then switched to the desk itself, then to the room where such desks were, then to the entire building with rooms where these desks are, and now it can mean “department”, many buildings with such rooms, some are not even in the vicinity. Same with managers, and AI should be the first one to spot the waste.
Translation – one can always hire back the managers needed. (as a side note, where I work it seems that they specifically seek the lowest possible IQ to hire the highest-paid managers – loyalty is paramount, technical expertise/knowledge is optional; at this rate AI can probably take over all the important tasks, and they won’t even know about it).
As Cory Doctorow always says, AI cannot take your job. AI doesn’t do anything at all unless someone makes it. Your boss can take away your job, because an AI sales-creep convinced them that an AI can do it. It’s important to make sure the actual responsible parties are named.
+1
This obsession with induced market destabilizations by AI evangelists is one of many red flags signalling the levels of dishonesty at play in representing AI capabilities. Monolithic expressions emerge in mediums which have great momentum, such as AI and CS. However, the most ubiquitous and likely realities of such a medium do not necessitate such a reverence. In fact, LLM research is merely a continuation of search engine research. Time has added complexity, no longer simply cross indexing data from webcrawlers, we add systems to work on natural language, and then DSP opens the door to any and all data (text, image, movie, audio). No use case of an LLM exceeds its inputs, and so these systems are taught to do that which AI was originally coded, to be a scrolling marquee. An electromechanical thing, which given human psychology we assume has faculty and intent and knowledge is simply letters shifting directionally.
Sending humans into space, with their heavy, high-list-of-requirements bodies, makes zero sense. If humans want to “go” to the stars, we’re going to have to send tiny simulations of ourselves or it’s not going to happen.
Yeah, I was going to say, I dont think I’d find being an astronaut particularly fun myself, on the contrary, a very stressful experience. It’s a bit like playing Call of Duty is fun, but being in a real war – not fun at all unless you were 100% psychopath.
“Taking jobs” as if it isn’t a new wave of human artists behind all these projects. You traditionalists have really got to get with the program or get out of the way.
This thinking comes from a place of fear, not love. AI cannot sustainably replace the jobs mentioned. I believe we would quickly notice the lack of human touch and want it back.
Meh, if the author knew anything, at all, about what goes into true flight control systems and aerospace communications, the title would be way different. AI takes the stupid jobs.