Oskar Höstlund

Swenglish husband & father of two | 'the snus guy' | 🇪🇺 | dev | washed up sub-elite runner | startups, defense, and finance | AI for a better life | 🇺🇦

[Oskar Höstlund @oskarhostlund]

Thanks for forging my fella @IballsInah!
Слава Україні!

[Trung Phan @TrungTPhan]

shoutout to Gary Larson for literally calling this in The Far Side in 1987 (!!!)

[Stefan Schubert @StefanFSchubert]

There are now more tech workers going from the US to Europe than the other way around

[James Jackson @derJamesJackson]

You can now bet on German train delays

[Oskar Höstlund @oskarhostlund]

The very first person to wish me happy birthday this year was my scrum master, uff.

[Oskar Höstlund @oskarhostlund]

Turn doubt to a payslip

[Oskar Höstlund @oskarhostlund]

@Brady_H @NoahRyanCo Germans meme that it is the worst one that we have though ahaha

[Fredrik Hjelm 🇸🇪 @FredrikHjelm4]

Why don’t more industries operate by the simple rule in tech?

"Be nice to everyone, especially the guy in the hoodie? He might build the future."

Esp bad in finance, politics. Where else?

[Oskar Höstlund @oskarhostlund]

Vibes don't come for free

[skooks @skooookum]

Every evening I go to bed excited to drink a coffee the next morning.

[Fredrik Hjelm 🇸🇪 @FredrikHjelm4]

And btw, most of the value lies in giving time back to humans

So they can focus on value creation, not in being human APIs

It’s not about replacing SaaS with another AI-coded app.

That’s a surface-level take.

The real value lies in orchestrating digital workers and agents across systems, owning the runtime, controlling how tasks are executed, monitored, and improved.

Some of it will look like an app to humans.

Most of it will run headless.

The goal is not cheaper software.

The goal is giving time back to humans.

[Ravid Shwartz Ziv @ziv_ravid]

The AI hype about replacing all of us has accelerated massively in the last few weeks.

But the main problem with AI predictions is that Nobody actually knows what AI will achieve.

Will the small bugs preventing end-to-end pipelines get solved? Will scaling keep working? No one knows.

But on Twitter, you get two types of people talking about it:

- People with very strong opinions who have no idea about the actual capabilities of future models. They just talk with confidence.

- And CEOs of OpenAI/Anthropic/xAI who do know more, but have so many incentives that it's hard to tell what's real and what's PR. (What about GPT-2 weights? Are they still dangerous?)

The honest answer is "I don't know." But that doesn't get engagement....

[vittorio @IterIntellectus]

the hardest filter in the modern world is the ability to sit with a problem for more than 15 minutes without checking your phone

that's it.
that's the entire competitive advantage now

[staysaasy @staysaasy]

Apple spending nothing on AI and watching people compulsively buy Mac minis, add tons of apps to the App Store, and everyone turning into a developer who needs a MacBook for their projects.

[Boring_Business @BoringBiz_]

This is why investing is such a brutal game

Go back 5 years to 2021. The world is shut down and everything is happening online

You could have been right about e-commerce. You could have been right about the cloud

And if you picked the company that is leading both of those categories, you got absolutely smoked by Walmart

[Meg McNulty @meggmcnulty]

@jameswise When you see this many deep-tech bets in one place, it signals momentum. Capital, talent and urgency are converging around hard problems again.

[Oskar Höstlund @oskarhostlund]

@tekbog Do both make you even better, burst of yapping before retreating for a while to be a yearner

[Ionut «John» Burchi @burconsult]

Can we stop calling them skills and prompts and just call them what they actually are: spells and invocations?

[terminally onλine εngineer @tekbog]

if you want to be a good software engineer you have to be a yapper or a yearner

the greeks were right once again

[⟠Palis⟠🐍 @palis]

People who understand LLM technology at the technical level understand that at the end of the day, its usefulness is in completion speed of well defined tasks leveraging existing information

This is incredibly useful and not to be downplayed as a technical tool, as information becomes better structured and tasks more well defined (also with the help of the speed of AI) almost everything becomes easier

But AI cannot think in any real sense, or create in any real sense, or experience, especially not any human experience. It can appear to do these things, and savvy people can easily fool unsophisticated people with an illusion it can do these things, but people aware of the technical fundamentals understand it cannot, and is still just a computer. You will still need engineers, you will still need designers, you will still need most technical roles

[Oskar Höstlund @oskarhostlund]

@Jenkins9odin9 That's awful, people are monsters