Testify

@PeterDiamandis Starting a company was never the hard part. Keeping it alive is...

Here's a truth: AI has made starting companies EASIER.

Nuclear power is an incredible technology. But once trust was lost, adoption slowed for decades. AI could face the same fate. The real bottleneck isn’t necessarily capability. It’s whether people trust the technology enough. Will people trust AI fully? Should they trust AI fully?

Don’t think in terms of white-collar vs blue-collar jobs being replaced by AI Think in terms of repetitive tasks. AI will automate tasks that are predictable and repeatable, no matter the profession. Stop doing on repetitive work and start focusing on harder, more interesting problems.

@Dostoevskyquot Masterpiece

what is your take on this book

You’ve probably seen this chart everywhere on social media. Before taking it at face value, ask a few simple questions: • What does the company that published it actually sell? In this case: technology that should replace human work. Therefore they will only publish something that helps them sell this idea • Can you really reduce someone’s job to a list of isolated tasks? If some tasks get automated, that often frees time for others...innovation, research, problem solving. • And remember: theory and practice are very different. Just because something is theoretically possible doesn’t mean it’s practical at scale. Real adoption is usually far messier than the charts suggest. AI will absolutely change how we work. But it’s still worth thinking critically and applying a bit of logic before accepting every viral graph.

@TheRundownAI Worth keeping in mind that this is a survey and chart published by the very company selling the replacement. That doesn’t mean change isn’t coming. But it does mean the source might have a slight incentive to frame the story a certain way.

Anthropic just studied which jobs AI can theoretically replace vs. which ones it's actually automating right now. Computer & math: 94% exposed. Legal: ~90%. Management, architecture, arts & media: all 60%+. Observed usage so far? A fraction of that. But the gap is closing fast. Every field where the blue line towers over the red is borrowed time. Grounds maintenance and construction are sitting at near-zero on both. Might be a good year to learn landscaping!

Daily reminder Especially with all the “AI layoffs” headlines

If writing code was enough, every developer would ship great software. If writing words was enough, everyone would be a great author.

Claude is down. And suddenly productivity drops everywhere. Dependence on AI is becoming a real risk.

@svpino Imagine if the goal of AI was augmentation, not replacement…

I love everything Anthropic is building and how fast they are shipping killer features. I hate everything Anthropic is doing to alienate developers, and I hate their hypocrisy. I hope they figure this out.

AI doom is everywhere on X lately. So I decided to write more about AI safety and alignment. AI is coming into the workforce. The real question is how we introduce it in a way that makes sense and actually benefits humans. The future of AI is not fixed (even though people think it is). We can still influence how it is built and how we live and work with it. Follow if you're interested in these topics, and let me know which questions and areas you want me to explore most.

@svpino It works until it doesn’t. If the code breaks and the AI can’t fix it and the human can not read the code, we have a real problem. And depending on how critical the system is, that’s a very dangerous place to be.

When building agents, we care about two things: intelligence and cost. We want to maximize intelligence while minimizing the cost of producing it. Should agents even produce human-readable code? Human-readable code is an intermediate layer that's only useful if humans must stay in the loop. But what happens when agents become truly competent? At that point, humans will become a bottleneck. Future agents may not write Python or JavaScript, or any other modern programming language. They may generate something entirely new, maybe even closer to bytecode. It will be unreadable to us, but much more efficient for them.

@benedictk__ Hope we’ll see you back in Zurich again soon 🇨🇭 There’s a lot happening in deep tech, robotics and AI, and the ecosystem keeps getting stronger every year.

Munich and Zurich are popping. Just got back to Vienna - beautiful place - but it feels like there is no tech or excitement at all.

@JonhernandezIA Maybe the real innovation in AI right now is knowing when to slow down. If it isn’t safe, it isn’t ready.

📁 Demis Hassabis, CEO of Google DeepMind, says AI could transform health and energy. But two dangers grow with it. People using it for harm. And systems becoming too autonomous to control. We may solve it. But we need to solve it fast.

Breaking: AI companies upset that other AI companies are scraping data that was scraped from scraped data. #anthropic #deepseek #ai #artificialintelligence

After using AI quite a lot, here’s the most useful skill for navigating AI no one talks about: Patience!!! I’m always one prompt away from just doing it myself

@AnthropicAI This might be the most unintentionally ironic take I’ve seen today…

We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax. These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.

@siddharthkp Everything is getting „killed“ and being „dead“ nowadays

this week on twitter, i've read: saas is dead UI is dead programming is dead what else is dead?

@damianplayer Most people don’t get paid just for what they know. They get paid for how well they apply it.

I just heard a Goldman Sachs exec say knowledge work is going to zero. and it’s something I’ve been thinking about.. every lawyer, consultant, accountant. they all get paid for the same reason. they know something you don't. that's it. that's the whole business model. AI made knowledge infinite. so what happens when everyone has the same information at their fingertips? what's actually worth paying for anymore?

Right now, “built with AI” seems to sell better than the product itself…