Tech Exec, Founder & Consultant • Advocating thoughtful AI Adoption
Cologne, Germany
Joined January 2009
- Tweets693
- Following217
- Followers228
- Likes178
💡 Fun fact Friday: Alphabet stock plummets after it announces the world's most capable model.
Why? First-ever negative free cash flow and AI Search cannibalising Ad revenue.
Markets think that borrowing money to invest in a technology that threatens your revenue is dumb.
9
After critical vulnerabilities were reported, our PHP friends will need the extra coffee today: they need to patch 70% of the world's websites, and quickly.
If your agency isn't bugging you about upgrades, that's on them.
If you are not listening, it is on you.
18
Merging Tesla and SpaceX sounds like a ‘good idea’ as the AI singularity approaches. A compelling argument for consolidation. I am swapping my shirt-and-trousers for a onesie, because, you know, singularity.
12
🟥 New post: "Expert (with AI), meet client (with AI)"
On a recent visit to a law firm, I was pleasantly surprised to meet a modern legal expert rather than a stuck-up lawyer.
humansafterall.tech/p/expert…
6
📉 Fun fact Friday: The J bends to an S. Anthropic falls short of expectations, as revenue growth appears to plateau.
On the face of it: Satisfaction = Performance - Expectation. But revenue is the only number that matters in AI.
1
32
"Create a website design for a digital widget." This is what the New Ugly looks like.
Algorithmic design will make things frictionless, usable, pleasant, beige.
As UI-slop floods browsers, it is time for designers to take the lead again.
24
🟥 New post: "Can you be trusted? Prove it"
As AI entrants flood the market, buyers discount claims of credibility. To earn their trust, businesses must demonstrate maturity and staying power.
humansafterall.tech/p/can-yo…
16
Fun fact Friday: Employment remains strong despite fears of AI-driven job displacement.
Democratisation of powerful tools does not just empower consumers. It also elevates the professionals who master them.
Data and chart from the FT: ft.com/content/8f783c04-5fc5…
16
Paradoxes live in the fault lines between logic and language.
How will AI adoption play out in the next few years? No one knows for sure. But these three frequently cited paradoxes might point us in the right direction.
Read the post: humansafterall.tech/p/ai-par…
1
11
When AI wreaks havoc, who is to blame? We need not overthink it: AI is not exempt from existing laws.
Flood the floor with lawyers and compliance suits, and the frontier will pace itself.
ft.com/content/2c24ece3-ac99…
7
🟥 AI is good at solutions and bad at decisions.
It operates in incomplete context and misses the interesting questions.
AI can save you time getting to an answer. But the cost of missed questions is much higher.
Read my latest post: humansafterall.tech/p/the-jo…
🤖 Made with AI
7
Time to "pace the frontier" was 6 months ago.
Back in March, I penned an open letter to Anthropic's CEO. I stand by my recommendations:
1. No autonomy without explainability
2. Restrict access to verified users
3. Force labs to accept liability
humansafterall.tech/p/an-ope…
1
14
“Never trust a computer you can’t throw out a window,” quipped Steve Wozniak.
Frontier AI labs are untrustworthy by design and are trying to lock themselves into your business. Owning a replaceable AI stack is easier than you might think.
humansafterall.tech/p/can-yo…
13
"Prod not God" is a great positioning statement by @typesafeai, but I will not shed my scepticism just yet.
True, if we were designing AI from scratch, it would likely look like Jed: cheap, accurate and designed as a component, not a product.
Let's see what happens next.
37
“There is no AI in Team”. I recently posted five predictions on the AI-enabled workplace of the future. But what will AI-enabled work actually feel like?
Roll forward to 2030, and I expect we will still moan and share memes after All-hands meetings.
humansafterall.substack.com/…
14
You're not truly "inspired" until you're asking yourself "what the f**k am I doing"
Mercedes Bidart's TED talk about financial inclusion and turning social signals into credit scoring data, was incredible. What a case study for AI.
#tedaivienna25
9
Day 2 at the TED AI conference in Vienna - and we are going into some deep questions.
On the nexus of AI and Design, how do we bring back friction ("productive resistance") to empower rather than dumb down our users?
14
Proud to see my fellow Europeans adding some finesse to the boorish and boastful tech industry.
The little red bridge is nice and all, but we'll be rocking AI in a 13th-century imperial palace.
One more sleep to TEDAI Vienna.
6
Soon, we will all be talking about "Composable Small Language Models".
A careful orchestration of SLM-based agents will likely drive Business AI Adoption. Why?
Easier to build and upgrade, cheaper and more energy-efficient to run.
economist.com/business/2025/…
11