Juhani Lehtimäki https://peertube.wtf/c/juhanilehtimaeki/videos My take on business, software, and the German industry where I have spent most of my career. I have been programming for thirty-five years, and getting paid for it for more than twenty-five. I started in Finland. I have been in Germany since 2008. I have been an engineer, a CTO, and until recently I ran a software company that sold into the German car industry. I have sat on both sides of that table. When I talk about how these companies work, I am not guessing. This is an opinion channel. Not journalism, not research. One thesis per video, stated plainly and defended. When I state a fact, it is sourced or I was there. When I am guessing, I say so. I never name clients. Not anonymously, not with a wink. If a story only exists because somebody trusted me with it, it stays off camera. The subject is never a person. It is the structure that makes intelligent people do expensive things. You may recognise the building. I also give conference talks. Business contact below. Sat, 24 Oct 2026 00:02:27 GMT https://validator.w3.org/feed/docs/rss2.html PeerTube - https://peertube.wtf Juhani Lehtimäki https://peertube.wtf/lazy-static/avatars/51e5f1a4-e86a-4c82-b806-1d08edce7f86.jpg https://peertube.wtf/c/juhanilehtimaeki/videos All rights reserved, unless otherwise specified in the terms specified at https://peertube.wtf/about and potential licenses granted by each content's rightholder. https://peertube.wtf/feeds/podcast/videos.xml?videoChannelId=44118 <![CDATA[Open Source Alone Won't Make Europe Digitally Sovereign]]> https://peertube.wtf/w/gS3GRQ1BgSfxD6U9z4AewC https://peertube.wtf/w/gS3GRQ1BgSfxD6U9z4AewC Fri, 23 Oct 2026 16:48:54 GMT Open source alone will not make Europe digitally sovereign. An open licence is not the same as open control. That is my take.

Your phone, your laptop, your email and your login all depend on American services. I think that dependency is a risk, and Europe needs to build its own answer.

The obvious answer is open source. I do not think it is a solution on its own. Some of it is not a product, and an open licence does not tell you who decides what goes in. You can build on open source, as long as it is real open source. You cannot just use it and call that sovereignty.

Copying Silicon Valley is not the answer either. Give it away free, lock people in with features they do not really need but still want, then squeeze. I would rather see products built for users, with business models people pay for because they want the software, not because they are stuck in it.

What I think we need is products people can leave. When a product goes bad, people should be able to take everything with them, including their followers. Protocols like Mastodon and ActivityPub already show that this can be done, even if it is mostly geeks for now. Identity and payments that are not tied to one business would be the next layer. Regulation can force interoperability. It will take time, and we have to start now. Something better, and something European.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software. Filmed on a riverbank.

CHAPTERS
0:00 Your phone, your laptop, your email
0:41 The trust is gone
1:20 Open source is not enough
1:58 Not another Silicon Valley
3:06 A fair playground
3:50 Products people can leave

If you work on this from the inside and think I got it wrong, tell me in the comments. I would rather be corrected than agreed with.

That is my take.

Music: StreamBeats by Harris Heller

]]> Juhani Lehtimäki nonadult Open Source Alone Won't Make Europe Digitally Sovereign Open source alone will not make Europe digitally sovereign. An open licence is not the same as open control. That is my take. Your phone, your laptop, your email and your login all depend on American services. I think that dependency is a risk, a... <![CDATA[Why Developers Are Fighting About LLMs]]> https://peertube.wtf/w/5bi37BqfG1EG4XdTXJP9uc https://peertube.wtf/w/5bi37BqfG1EG4XdTXJP9uc Fri, 23 Oct 2026 16:48:43 GMT The software industry is more divided than I have ever seen it. I lived through PC versus Amiga and Android versus iOS. That was passion. This is hatred.

My take: a big part of the LLM fight comes from calling two different jobs by the same name. Product engineers solve problems for real users. Software engineers solve problems in the code. For the first job, these tools have already changed everything: you do not have to see the code anymore. That is a painful thing to hear if you spent your career learning every detail of it.

I do not think LLMs are going anywhere, whatever you call them. I still refuse to call them AI. They have real problems, and I say so: the environmental cost, the way the American hyperscalers operate, copyright. They are also productive, and productive tools get used. I care about the environment. I do not even own a car. And I use LLMs a lot. I am probably a hypocrite, and I would rather say that myself.

Nobody knows where this industry is heading. Me included. But the world keeps spinning, and we will figure this out.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software. Filmed walking an unused railway line.

CHAPTERS
0:00 The software industry is in chaos
0:23 Change is coming
0:49 Not just another flamewar
1:29 Horse versus car
2:17 Two jobs, one title
5:54 Slop or purpose
7:18 Where I struggle
8:28 What now

If you are on the other side of this argument, tell me in the comments where I got it wrong. I would rather be corrected than agreed with.

That is my take.

Music: StreamBeats by Harris Heller

]]> Juhani Lehtimäki nonadult Why Developers Are Fighting About LLMs The software industry is more divided than I have ever seen it. I lived through PC versus Amiga and Android versus iOS. That was passion. This is hatred. My take: a big part of the LLM fight comes from calling two different jobs by the same name.... <![CDATA[How German Car Industry Destroyed Itself]]> https://peertube.wtf/w/fpAHAXRJSL4hW2UJzs29od https://peertube.wtf/w/fpAHAXRJSL4hW2UJzs29od Fri, 23 Oct 2026 16:48:24 GMT The German car industry did not lose because somebody built a better car. The cars were never the problem. The problem was the decisions.

Decisions a competitor makes in a week took these companies years. Some never got made at all.

I have spent 25 years in the software industry and 15 of them running my own companies. In recent years a lot of that time went into building in-car software and selling to the German car industry. It is an unusual seat. You are not inside any one of these companies, which means you can compare them: against each other, and against every other industry you have worked in. When you compare, you see they are all making the same mistakes. Not similar mistakes. The same ones.

This is my take on how it happened, in three parts. They could not decide. They could not build software. And they spent ten years suppressing demand for the product they were going to have to sell.

None of it needed anybody to be stupid. That is the uncomfortable part, and it is why this happens to incumbents everywhere, not just in cars.

I am not a neutral observer, and this channel is my opinion, based on what I have actually seen. Where the video names a company, it is standing on public reporting rather than on anything I was told in private. Those reports are linked below so you can check them yourself.

CHAPTERS
00:00 The cars were never the problem
00:46 Who I am, and why I am not neutral
01:07 They were extraordinarily good at this
01:41 One: they could not decide
03:08 Two: they thought software was a part
04:08 The last line on the spreadsheet
07:09 The promotion algorithm
08:30 The receipt: CARIAD
09:31 The admission: buying software from Rivian
10:24 Three: they killed their own home market
11:10 The lobbying worked, and that was the problem
12:24 The fair counterargument
12:48 Nobody decides to lose

]]> Juhani Lehtimäki nonadult How German Car Industry Destroyed Itself The German car industry did not lose because somebody built a better car. The cars were never the problem. The problem was the decisions. Decisions a competitor makes in a week took these companies years. Some never got made at all. I have spent... <![CDATA[What we need is products people can leave]]> https://peertube.wtf/w/4CW3e8WY96Jsbb3Dhww7bt https://peertube.wtf/w/4CW3e8WY96Jsbb3Dhww7bt Fri, 23 Oct 2026 16:48:02 GMT What we need is products that people can leave. When a product is bad, people should be able to leave and take everything with them, including their followers.

That is not a fantasy. There are protocols that let you take your followers with you when you leave a social media platform. Mastodon and ActivityPub are the examples. It is mostly geeks and nerds for now, but it proves it can be done.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult What we need is products people can leave What we need is products that people can leave. When a product is bad, people should be able to leave and take everything with them, including their followers. That is not a fantasy. There are protocols that let you take your followers with you w... <![CDATA[The EU has a lot of problems. It can still help.]]> https://peertube.wtf/w/eCwbgcafgvRSbhvQ2e3s5z https://peertube.wtf/w/eCwbgcafgvRSbhvQ2e3s5z Fri, 23 Oct 2026 16:47:57 GMT The European Union has a lot of problems. It is also a rare place where we have a governing body that can actually help.

I want Europe to have a healthy, competing ecosystem and a fair playground: products people actually want to use, and business models people pay for because they want the software, not because they are stuck in it. The EU can help create a playground where products compete, not filthy business models.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult The EU has a lot of problems. It can still help. The European Union has a lot of problems. It is also a rare place where we have a governing body that can actually help. I want Europe to have a healthy, competing ecosystem and a fair playground: products people actually want to use, and busines... <![CDATA[What's the point if we just build another Silicon Valley?]]> https://peertube.wtf/w/5SCr8FmVhL6jocv6p1Ztpy https://peertube.wtf/w/5SCr8FmVhL6jocv6p1Ztpy Fri, 23 Oct 2026 16:47:51 GMT It will take time to get this right, but we have to do it now. We have to build products, operating systems and apps that are not using the Silicon Valley model.

What is the point of doing this if we just build another Silicon Valley? We want to do something better. We want to do something European.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult What's the point if we just build another Silicon Valley? It will take time to get this right, but we have to do it now. We have to build products, operating systems and apps that are not using the Silicon Valley model. What is the point of doing this if we just build another Silicon Valley? We want to ... <![CDATA[I'm probably a hypocrite]]> https://peertube.wtf/w/7eb9m7g6THQPAvoFaA6PaJ https://peertube.wtf/w/7eb9m7g6THQPAvoFaA6PaJ Fri, 23 Oct 2026 16:47:45 GMT I don't own a car. I ride a bicycle everywhere, and I tell people not to drive because cars ruin the environment. I also use LLMs a lot, and I've been told not to, for the same reason.

I have no answer to this. I'm probably a hypocrite, and I would rather say that myself.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult I'm probably a hypocrite I don't own a car. I ride a bicycle everywhere, and I tell people not to drive because cars ruin the environment. I also use LLMs a lot, and I've been told not to, for the same reason. I have no answer to this. I'm probably a hypocrite, and I wou... <![CDATA[You don't have to see the code]]> https://peertube.wtf/w/3pUBXq6XY2hFyNawEEzK8N https://peertube.wtf/w/3pUBXq6XY2hFyNawEEzK8N Fri, 23 Oct 2026 16:47:39 GMT When you are building products for users, you do not have to see the code anymore. That is a painful realisation if you spent your career learning every nuance of it.

When I went to university in the 1990s, I was told Java would never become popular, because only a human knows how to manage memory. That turned out not to be true. It is not quite the same, but it is similar.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult You don't have to see the code When you are building products for users, you do not have to see the code anymore. That is a painful realisation if you spent your career learning every nuance of it. When I went to university in the 1990s, I was told Java would never become popu... <![CDATA[Two jobs. One title.]]> https://peertube.wtf/w/59JGkMGifRF4ER1i179zE5 https://peertube.wtf/w/59JGkMGifRF4ER1i179zE5 Fri, 23 Oct 2026 16:47:30 GMT Product engineers solve problems for real people. Software engineers solve technical problems in the code: performance, security, safety. We call both of them software engineers, and I think a big part of the LLM fight comes from that.

For product engineers, LLMs are super powerful. Software that used to take weeks or months now takes days, tested with real users.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult Two jobs. One title. Product engineers solve problems for real people. Software engineers solve technical problems in the code: performance, security, safety. We call both of them software engineers, and I think a big part of the LLM fight comes from that. For produc... <![CDATA[Now there's hatred]]> https://peertube.wtf/w/7U5Vn6WMinu3AVCFj4QBim https://peertube.wtf/w/7U5Vn6WMinu3AVCFj4QBim Fri, 23 Oct 2026 16:47:20 GMT I lived through PC versus Amiga and Android versus iOS. Those were flame wars, but it was passion. Never hatred. The LLM fight is different.

Whatever you think about LLMs, they have proven to be extremely valuable tools for building products. What you think about them doesn't really matter. They are coming.

This channel is my opinion, based on what I have actually seen in more than twenty-five years in software.

]]> Juhani Lehtimäki nonadult Now there's hatred I lived through PC versus Amiga and Android versus iOS. Those were flame wars, but it was passion. Never hatred. The LLM fight is different. Whatever you think about LLMs, they have proven to be extremely valuable tools for building products. Wha... <![CDATA[The Real EV Disinformation]]> https://peertube.wtf/w/hQsm6xTiAHp2Gqv7jF6oDL https://peertube.wtf/w/hQsm6xTiAHp2Gqv7jF6oDL Fri, 23 Oct 2026 16:47:09 GMT The biggest disinformation about electric vehicles in Europe
hasn't come from the oil companies. It's come from the car
manufacturers.

The market isn't ready. Customers don't want them. The range
isn't there. The grid can't take it. What they actually wanted
was one more round of combustion vehicles — because a combustion
car sold today is margin today, and investment in EVs is margin
later. Bonuses are tied to quarterly and yearly results.

So this wasn't a failure. They succeeded. German car companies
built enormous political power and spent it lobbying against
electric vehicles. And now they're done. They killed their home
market.

#ev #electricvehicles #carindustry #germany #automotive

]]> Juhani Lehtimäki nonadult The Real EV Disinformation The biggest disinformation about electric vehicles in Europe hasn't come from the oil companies. It's come from the car manufacturers. The market isn't ready. Customers don't want them. The range isn't there. The grid can't take it. What they act... <![CDATA[Software Engineers Do Not Stack]]> https://peertube.wtf/w/jYLdkzJimTYdzWZMVz1EpP https://peertube.wtf/w/jYLdkzJimTYdzWZMVz1EpP Fri, 23 Oct 2026 16:46:58 GMT It matters who writes the software. You need a few good engineers
— not a lot of mediocre ones.

A factory gets more cars out by adding a night shift. Software
works the exact opposite way. Communication overhead grows faster
than the number of people you add, so past a certain point every
engineer you add to speed things up slows the project down. Every
team you add slows it down.

35 years writing software, 25 in the industry. This one is not
controversial among people who build things — and it's still the
first assumption a purchasing process breaks.

#ev #electricvehicles #carindustry #germany #automotive

]]> Juhani Lehtimäki nonadult Software Engineers Do Not Stack It matters who writes the software. You need a few good engineers — not a lot of mediocre ones. A factory gets more cars out by adding a night shift. Software works the exact opposite way. Communication overhead grows faster than the number of pe... <![CDATA[Nobody Gets Fired For Asking For More Data]]> https://peertube.wtf/w/2j3ZVSYQSAJGgtbpJ27XQ5 https://peertube.wtf/w/2j3ZVSYQSAJGgtbpJ27XQ5 Fri, 23 Oct 2026 16:45:46 GMT Not deciding doesn't feel like a mistake. It feels like being
careful. It isn't. It's deciding to arrive late.

Everyone in the room already knows how it should be done. But
instead of deciding, someone asks for more information. A
committee gets triggered. A workgroup. More documents. By the
time a recommendation arrives, the market has moved and the
evaluation is invalid — so they run another one.

Nobody is incompetent here. Everybody is doing their job right.
Make a wrong call and your name is attached to a failure. Ask for
more data and you're just being diligent.

I've sat in many meetings whose only outcome was the timing of
the next meeting.

#ev #electricvehicles #carindustry #germany #automotive

]]> Juhani Lehtimäki nonadult Nobody Gets Fired For Asking For More Data Not deciding doesn't feel like a mistake. It feels like being careful. It isn't. It's deciding to arrive late. Everyone in the room already knows how it should be done. But instead of deciding, someone asks for more information. A committee gets ... <![CDATA[Cars Were Never The Problem]]> https://peertube.wtf/w/drjW8CFbeGDsmG7gbECScG https://peertube.wtf/w/drjW8CFbeGDsmG7gbECScG Fri, 23 Oct 2026 16:44:54 GMT The German car industry didn't get destroyed because somebody
else built a better car. It was always decision making.

I've spent years building in-car software and selling it to these
companies. From the outside in, you can compare them — and they
don't make similar mistakes. They make the same mistakes.

The reason is simple: in Germany, people rotate through these
companies. A platform lead at one is the platform lead at a
competitor next week. Same people, same mistakes.

Full video: https://www.youtube.com/watch?v=40MJ6-xQvvI

#carindustry #softwareengineering #germany #automotive #tech

]]> Juhani Lehtimäki nonadult Cars Were Never The Problem The German car industry didn't get destroyed because somebody else built a better car. It was always decision making. I've spent years building in-car software and selling it to these companies. From the outside in, you can compare them — and the...