tante (his nickname is lowercase!) caught my attention on Bluesky with his excellent critiques of Gen AI and other Silicon Valley bullshit, such as blockchain and NFTs.

He really knows his stuff when it comes to code, software engineering, and how that relates to society and culture. From the bio on his website:

tante/Jürgen Geuter studied computer science and philosophy at the Carl von Ossietzky University of Oldenburg and worked there for a while as a researcher and teacher.

After working for a few years as project manager and consultant for a software company working on industrial maintenance and production systems, he now works as the Research Director for the spatial new media studio ART+COM in Berlin.

Besides his work in and with tech he’s been a writer and freelance consultant since the beginning of the 2000s writing about the social, political and structural impact of the way tech is implemented, conceptualized and talked about starting with issues around privacy and data usage and work on “AI”. In 2018 he was a founding member of the interdisciplinary Otherwise Network.

A true hacker would never like Gen AI.

I asked tante some questions over email.

Kim Crawley: What was your professional journey into the world of computer technology?

tante: I grew up fiddling around with computers (mostly changing the operating system settings and startup files to leave enough memory to run games), skills that also got useful when doing layouts for an independent school paper and small websites while I was in school. I then went to study Computer Science with a minor in Philosophy, which I finished (after more years that I should probably admit). I finished with a German Diploma (which is basically a Master’s degree). While still a student I made a living developing software (either for websites or all kinds of backend services) and got into Linux so I also did a lot of Linux server administration jobs.

After a few years teaching at the university, I graduated from I got a job doing Industry Automation (consulting, developing and project management), which went on for a few years. That job gave me a lot of insight into how management sees automation and how those automation systems were often being developed against the workforce and their explicit recommendations. Technically it was all very interesting, the combination of big hardware (think huge factories and all the robots and machines in them) and software running those shop floors was challenging. But I also developed a new understanding for the non-technical aspects of IT: The need for doing things right (because otherwise people might lose an arm) and also the understanding that every IT project is a social project first and a tech project maybe second, probably third. It was a very useful reality check.

I then went on to work for a company that builds interactive experience spaces (think museums, media art installations, visitor centers for bigger companies) where I am currently running the research department. In that job I am trying to find funding to experiment with new (and old) technologies to find new modes of interaction that go beyond what people carry in their phones and that do support learning through cyberphysical experiences. This is less about “optimization” but about making the visions and abstract concepts of designers reality or giving those designers new technologies or new setups of old technologies to play with.

Aside from my day jobs, I have always kept some freelance/pro bono work in software development/administration going to stay in touch with what’s going on.

These days I don’t get to build things myself too much (due to lack of time), today it’s more about giving people a few hopefully inspiring and/or dangerous ideas and providing the funding they need to explore things.

Crawley: Why do some of us become skeptics of Silicon Valley culture and marketing, while others drink the Kool Aid and jump on blockchain, NFT, Gen AI bandwagons?

tante: I cannot speak for others, and I do think that there are very individual motivations for not going all in on whatever Silicon Valley is selling. So I’ll focus on my own thinking here. I think it’s a few things coming together.

The first thing is this realization that with all the claims of world-changing technologies, nothing “better” has come from Silicon Valley in like 15 to 20 years. Everything is magic and a revolution, but software gets worse and slower with every iteration to the point where people actively do not update their tools to avoid them turning into garbage in their hands.

I am an engineer, and I feel like software engineers /computer scientists (which at least in Germany, do consider themselves to be engineers) do have an obligation to provide good, secure, high quality products and solutions to society. Engineering is not a job, but a profession in my understanding and with that comes responsibility. We should be pushing for better quality. It’s ridiculous how you need immensely powerful processors these days to run text editors or other apps without stuttering. We dropped the ball and Silicon Valley/Startup Culture was the fuel that really turned that fire into what we have now.

I also come from a very requirements/people centered perspective. I don’t think that thinking about technologies without context or in “potential” is all that useful. “What is this for?”, “What problem is this solving and at what cost?”, “Is this really the best solution?” are questions that guide my thinking about technology way more that whatever magic some corporation or middle-aged duded on LinkedIn might push.

So in the end for me, it is largely about what systems actually provide. And with many of the recent hype technologies that equation never made sense: I still don’t think that there is a legitimate use case for blockchains in the real world that would be solved better with other tools/technologies. I am also very skeptical of the actual value of stochastic pattern generators (like) “Gen AI.” Because. how often is your requirement: “I need to quickly generate some text that looks good, but might be utter garbage, so I have to spend a lot of time editing and checking a string of characters that might not make any sense.”

That has always been my criticism. Silicon Valley pushes these “the future” narratives that might look cool when you just glance at them from afar. But when you look at the technologies, understand them for what they are built for, what the tradeoffs and capabilities are, a lot of the Silicon Valley pitches don’t make much sense. Either technologically or socially or economically. And often it’s not “the tech being bad” (except maybe for blockchains) it’s just that tech is being used in ways that don’t work in this reality we live in. So why are we talking about them?

We have so many problems to solve, why are we talking about whether an LLM will kill us all if we are not nice enough to it? That is at best a sign of avoidance, and at worst a sign we all need to have our heads and hearts examined.

Crawley: One of my gigs now is I have to co-host a live video show on LinkedIn that’s sponsored by an American cybersecurity vendor. My boss is a clanker (derogatory term for someone who loves Gen AI), my co-host is a clanker, our interview subjects are clankers, and I’m paid just about enough to bite my tongue and just occasionally say things like “I actually do all my own research and writing, as I did in 2018.”

Do you have any advice for me?

tante: Life’s complicated. My job entails writing grant applications for us to get funding to get to experiment. Currently in Germany where I live you only get money if you claim to do “AI” somehow. So I wrote a bunch of proposals for supposed “AI” projects. At the end of the day we all need to pay some bills and living off of just righteous anger is quite hard. So keep being a person who sometimes throws a wrench into the Gen AI conversation (until in a few years it all goes to shit and you take your victory lap). (Editor’s note: tante could write machine learning protect proposals as “AI.” The fools don’t know the difference!)

But in all seriousness. There is something beyond mere pragmatism and the need to feed and clothe oneself and one’s family. I do think that being in those kinds of environments is harmful to us as human beings, to our souls. Being forced to work in these environments can be very draining and can even sour one’s appreciation of things we might actually like (I do still like and enjoy working with tech for example).

It is a struggle that many people share, and in the end I only see two good options. If possible, find allies in your organization or context that allows you to push back, join a union, etc. But that mostly works for bigger corporations. Otherwise, it’s hopefully possible to find ways of living a decent life without having to be tied to those kinds of hype. For some that might be joining NGOs and working there, some might create their own thing. But as someone who is himself trying to find a way to spend my days avoiding the psychic damage of another person clueless about the actual properties and capabilities of so called “AI” telling me how efficient it will make everything I hope you find a way and show other how it might work.

Crawley: I was teaching enterprise cybersecurity at the Open Institute of Technology and I recently quit due to a combination of being underpaid and being forced to use TurnItIn e-proctoring rather than being trusted to be able to detect when students are using Gen AI with my own brain. Do you hear of many people in our professions with similar problems now?

tante: Yes, in many areas. Vibe-Coding, Vibe-Writing, Vibe-anything is super taking over in many areas with a massive push of “well you better join, because otherwise we’ll kick you out.” Like, I have seen highly paid consultants ask ChatGPT for solutions in front of their clients. What’s the strategy here? Why should anyone pay you your 4 digit daily rate, if your expertise can be rented for 20 bucks from Open AI?

Brian Merchant has been collecting a lot of stories from people whose job was “taken by AI” (AI of course never takes jobs but CEOs thinking that quality does not matter do). It’s kinda everywhere. We’re living in a time where there is almost an open war against competence. Because competence wants to be paid. And CEOs can’t have that. (Brian Merchant’s AI Killed My Job series is here.)

Crawley: How would you explain the problems of Gen AI to a layperson?

tante: It really depends a bit on the context, of course. There are so many problems, and not all affect everyone the same. But I try to make things very tangible. Think about output quality. Every LLM will add an asterisk to their output basically saying “you cannot trust anything I generated, you have to check it.” Which we are supposed to accept. But why? If I bought some milk at the supermarket and they attached a sticker to it “Might be full of rat poison, you better check” we wouldn’t accept that at all. We’d sue their asses.

I think that making the often abstract damages to our ecological, social, economical, psychological world more tangible often works. People are supposed to be using ChatGPT for therapy? Okay, should we allow any random dude to give people therapy? No? So why do we let a stochastic parrot that has not qualifications do it?

Crawley: Have experiences in your work informed you of how Gen AI is toxic?

tante: Not really, mostly cause Germany is always a bit slow with everything so me being interested in things makes me see issues often a bit before the rest of the company jumps on the hype. Some patterns have confirmed my impressions of toxicity and the way “Gen AI” is used against workers and their competence. People no longer being able to actually explain what their problem is beyond “it doesn’t work,” people without experience presenting half-baked “solutions” that don’t work with the understanding that someone “just needs to fix it,” when you know that fixing that hot mess of generated code will take longer than just writing it from scratch.

Crawley: How can readers learn more about your work and support you?

tante: If you want to follow my “work” it’s mostly, on my own website: tante.cc. I do troll the weird people on LinkedIn, and am active on the Fediverse (@tante@tldr.nettime.org, the instance he’s the owner of!), and a bit on Bluesky (@tante.cc).

I am trying to reduce my reliance on a “day job” a bit to put more focus on my work so people can of course throw a few bucks my way via a Patreon, which doesn’t really give you any extras though. I don’t want to lock my work behind paywalls, so it’s more a donation box. It’s not really doing much yet, but at least it pays for my internet bills.

Crawley: Is there anything else you’d like to add?

tante: These technological hypes can be overwhelming. I might feel impossible to do anything. And maybe it is to a certain degree. But I have made the experience that people need to hear something more often so here goes:

The thing you feel in your bones, the feeling that all these hyped techs are often bullshit, and don’t always make much sense is valid. Even if everyone seems to disagree. Don’t let people discourage you from trusting your gut. Have honest conversations with your peers. Find friends to talk these things through with, even friends who disagree with you. Build or find spaces where those conversations can happen respectfully. Either you’ll learn that you were wrong or you’ll convince others. Modern tech (as an appendage of capitalism) is focused on isolating you, keeping you separated from your peers and the rest of humanity. Don’t let that happen.

Oh, and do something creative, non techie that you are totally shit at. Do crochet. Learn to play the drums. It’s a great antidote because it connects you to the world, your body in it and other people. Doing something tangible, real with your body just for your own enjoyment and fun is the most radical act you can take.

---

If you are an activist, academic, technologist, writer, or artist who is bringing attention to how fucked up Gen AI is and you’d like to be interviewed for this series, email Kim Crawley at kim.crawley (at) stopgenai.com. I will definitely link to and plug your work if you’re cool!