ventures

Most people still think learning AI means:
→ buying another $500 course
→ watching random tutorials at 2am
→ collecting certificates they’ll never use
Meanwhile, some of the best AI education in the world is already online for free.
And honestly?
We’re entering a phase where the winners won’t necessarily be the people with the best degrees.
They’ll be the people who learn fastest and apply fastest.
Harvard, CS50, DeepLearning.AI and other institutions are quietly making world-class AI education accessible to everyone.
No gatekeeping.
No expensive bootcamp.
No application process.
Just real material covering:
→ Generative AI fundamentals
→ Prompt engineering
→ RAG systems
→ AI agents
→ LLM architecture
→ AI in education and business
→ Practical implementation
The information gap is disappearing.
The execution gap is becoming everything.
And that changes the game completely.
Because for the first time in history, someone with:
→ a laptop
→ internet access
→ curiosity
→ consistency
…can learn the same foundational concepts as people inside elite programs.
For free.
But here’s the uncomfortable truth:
Most people still won’t do it.
Not because the resources are missing.
Because passive learning feels productive while changing nothing.
People binge AI content for months and still never build:
→ a workflow
→ an automation
→ a prototype
→ a system
→ a business
Knowledge without implementation becomes entertainment.
The people who benefit most from AI won’t necessarily be AI researchers.
They’ll be operators.
The marketers.
The designers.
The founders.
The consultants.
The educators.
The small business owners.
The people who learn just enough to redesign how they work.
And this is where most professionals are underestimating the shift.
AI is not becoming a niche technical skill anymore.
It’s becoming operational literacy.
Like:
→ using Google in the 2000s
→ Excel in finance
→ social media in marketing
At some point, it stops being “specialized.”
It becomes expected.
My advice?
Don’t try to consume everything.
Pick one area:
→ prompting
→ AI agents
→ automations
→ RAG
→ AI for marketing
→ AI for business operations
Then:
→ learn
→ build
→ apply
→ repeat
That loop matters more than any certificate ever will.
Because the biggest advantage right now is not access to AI.
Everyone has access.
The advantage is learning how to think with it before everyone else does.

Переписать
Approved

How Codex can change your life?

The moment AI starts operating your computer instead of just answering questions, the whole game changes.
You can now literally say:
“Find this GitHub repo. Deploy it on my server where my other tools already run. Test it. Send me the result.”
And Codex can actually do it.
Or:
“Open my desktop. There are thousands of screenshots there. Create folders and organize everything logically.”
Done.
That’s not prompting anymore. That’s delegation.

A few things became very obvious to me while working with these systems.

First:
Talking to agents by voice is often much better than writing prompts.
When people type, they try too hard to sound smart and structured.

But agents often need your real chaos:
“I honestly don’t know what’s happening, everything is a mess.”
That’s useful context.

Second:
When the agent asks you to do something manually, push back.
“Do it yourself.”

Most people still operate with old reflexes and take control too early, while the agent can already handle huge parts of the workflow itself.

Third:
File structure suddenly matters again.
Folders, naming, context, documentation.
Because the agent knows nothing unless your environment is structured clearly.

And the biggest realization:
AI is splitting people into two groups.
The first group has vision.

They understand products, people, positioning, meaning.
But they don’t fully master every new AI workflow yet.

The second group mastered every AI stack imaginable.
Agents. MCPs. Automations. Claude workflows.

They build insanely fast.
But many quietly don’t know what actually deserves to exist.
And honestly, I’m 100% in the first group.

I don’t keep up with every new AI tool anymore.
And I don’t even think it makes sense to try.

New frameworks, agents, wrappers, MCP stacks, workflows appear every week. It’s an endless sandbox where nobody fully “wins”.

I’m much more interested in:
– what should exist
– why people would care
– what has emotional gravity
– what actually deserves to be built
That’s where I’m strong.

So if you’re in the second group – technical, fast, deep in tools, automations, agents – but feel like your product still lacks clarity, positioning, narrative, meaning, human understanding or product intuition…
Come talk to me.

Seriously.

Because right now the most dangerous combination in tech is not “more AI”.

It’s people with systems + people with vision.

That’s where real companies will come from.
1-hour sessions:
https://zcal.co/axlindholm/1hour
#AI #founders #product

• * 1
• What people see:
• – raised ~$12M across startups
• – Yes!Delft (twice)
• – living in The Hague
• – helping student startups go from idea to $70K+ ARR
• – building an accelerator and a venture boutique
• What people don’t see:
• – 60+ hour weeks (not occasionally… consistently)
• – 3 failed startups, some even with accelerator support
• – lost a loved business because of the war
• – rejected by 12 accelerators
• – years of nights of rethinking everything, over and over
• – 2 weeks living in a car after a forced and unjust divorce… just to keep going and paying my team
• – being a father and raising a child at the same time
• – navigating ADHD and overstress
• – migration pressure (including being told by authorities to “leave the country and raise your child via WhatsApp”)
• – legal pressure on what I can and cannot say publicly
• And still… here I am. With you and for you, guys.
• Funny thing – none of this fits into a braggy pitch deck.
• But this is the actual part of building.
• If you’re in that phase right now – tired, uncertain, still pushing…
• you’re probably closer than it feels.
• And yeah – keep an eye on what we’re building. Something real is coming!
• And if you’re going through something similar and need a more human conversation – not just “how to grow faster” but how to stay in the game – I keep a few free slots open 40 mins sessions:
• https://zcal.co/axlindholm/40min
• Take care of yourself. That part matters too.
• #founders #startups #reality

2

Everyone wants a “solo founder with AI agents”.
Reality is… less romantic.
Look at how co-founders of top companies actually met:
Mostly:
– worked together
– studied together
Not Twitter.
Not AI matching.
Real life.
And here’s the point:
Even now, startups are still built on people.
Not tools.
Most failures are not about tech.
They’re about people.
Wrong co-founder.
Or no co-founder at all.
Your environment is not “just people”.
It’s your future company.
Choose carefully.
Curious – solo founders or still a team game?
#startups #founders #AI

Extra
* 1
The biggest AI upgrade right now isn’t a new model.
It’s a shift in how we work with the ones we already have.
Anthropic just released Claude Skills 2.0.
And it quietly moves AI from “tool” to something closer to an operating system 🤖
Most people still use Claude like a chat.
Open → write prompt → tweak → repeat tomorrow.
Nothing wrong with that.
But it doesn’t compound.
The real shift happens when you stop prompting…
and start building Skills.
A Skill is simple:
One file that captures how you think and work.
Write it once, and Claude:
– applies your style automatically
– runs your workflow step by step
– follows your structure
– triggers when needed
No re-explaining.
No copy-paste loops.
Then 2.0 adds something interesting.
Claude doesn’t just execute your Skill.
It challenges it.
– it tests outputs automatically
– compares variations internally
– sometimes removing parts improves results
– rewrites triggers so they actually fire
Which is slightly uncomfortable 😅
but also powerful.
Because now AI isn’t just responding.
It’s refining how you work.
That’s the real shift:
From “better prompts”
to repeatable systems.
And once you feel that, going back to manual prompting feels… slow.
I recently put together a step-by-step guide on this:
– how to build your first Skill
– how to test it properly (most people skip this)
– the common failure modes
– how to turn Claude into a real workflow system
If you want it, just comment “Claude Skills” and I’ll share it 🤖
If you want help turning this into real workflows inside your startup or team, let’s go deeper:
https://calendly.com/inspirexchange/30min-crashtest
#AI #Startups #Productivity

• 2

Something is starting to move in South Holland.
Around The Hague, Delft, Rotterdam, Leiden.
And it’s a good signal.
More conversations.
More alignment.
Even institutions like CID are starting to see that mobility is not just about transport, but also social mobility. Who gets access. Who gets included.
At the same time, there is still a gap.
Many local players still don’t really know what is happening inside hubs.
For them, it’s not always clear why it matters or how to engage.
And maybe that’s on all of us.
Not to build things in parallel, but to be more open, more practical, and yes, a bit more proactive in working together.
Not events for each other.
Events with each other.
Founders.
VCs.
Hubs.
Municipalities.
Local businesses.
International talent.
Same room. Same conversation. Same direction.
Because in the end, the goal is simple:
Make the region more innovative, more connected, more economically strong, and more socially accessible.
I’ve been thinking a lot about how this could look in practice.
A few ideas we are exploring:
– a clear startup entry point for local and international founders, with real day-to-day support and onboarding
– VC tracks doing small, fast checks into validated local startups
– consistent weekly formats where people actually meet, work, and close things, not just talk
Less fragmentation.
More execution.
We are putting together a more structured plan around this and will share it soon.
For now, I’m genuinely curious to hear from others in the ecosystem.
If you are building, investing, running a hub, or working in policy, what do you feel is missing right now?
Feel free to share here or reach out directly if you want to exchange ideas:
https://zcal.co/axlindholm/1hour
NEW for MAY
NEW for MAY

Approved 1

How Codex can change your life?

The moment AI starts operating your computer instead of just answering questions, the whole game changes.
You can now literally say:
“Find this GitHub repo. Deploy it on my server where my other tools already run. Test it. Send me the result.”
And Codex can actually do it.
Or:
“Open my desktop. There are thousands of screenshots there. Create folders and organize everything logically.”
Done.
That’s not prompting anymore. That’s delegation.

A few things became very obvious to me while working with these systems.

First:
Talking to agents by voice is often much better than writing prompts.
When people type, they try too hard to sound smart and structured.

But agents often need your real chaos:
“I honestly don’t know what’s happening, everything is a mess.”
That’s useful context.

Second:
When the agent asks you to do something manually, push back.
“Do it yourself.”

Most people still operate with old reflexes and take control too early, while the agent can already handle huge parts of the workflow itself.

Third:
File structure suddenly matters again.
Folders, naming, context, documentation.
Because the agent knows nothing unless your environment is structured clearly.

And the biggest realization:
AI is splitting people into two groups.
The first group has vision.

They understand products, people, positioning, meaning.
But they don’t fully master every new AI workflow yet.

The second group mastered every AI stack imaginable.
Agents. MCPs. Automations. Claude workflows.

They build insanely fast.
But many quietly don’t know what actually deserves to exist.
And honestly, I’m 100% in the first group.

I don’t keep up with every new AI tool anymore.
And I don’t even think it makes sense to try.

New frameworks, agents, wrappers, MCP stacks, workflows appear every week. It’s an endless sandbox where nobody fully “wins”.

I’m much more interested in:
– what should exist
– why people would care
– what has emotional gravity
– what actually deserves to be built
That’s where I’m strong.

So if you’re in the second group – technical, fast, deep in tools, automations, agents – but feel like your product still lacks clarity, positioning, narrative, meaning, human understanding or product intuition…
Come talk to me.

Seriously.

Because right now the most dangerous combination in tech is not “more AI”.

It’s people with systems + people with vision.

That’s where real companies will come from.
1-hour sessions:
https://zcal.co/axlindholm/1hour
#AI #founders #product

Approved 2

Start Up Visa is not what you expected!
Just realized this recently: half of my active network in the Netherlands came here through the Startup Visa.
And their stories are wildly different.
A few built something huge.
Some run a small businesses.
Some never validated the original idea at all.
And some have been pivoting every few weeks for 5+ years 😅
But honestly?
With every iteration, their chances usually get better.
That’s the part many people outside the ecosystem don’t see.
Not everyone becomes a unicorn founder.
And that’s okay.
Some thoroughly built small European businesses.
Created jobs.
Helped local communities.
Found stability, freedom, meaning.
Behind every new Dutch B.V. there are real people maintaining years of their entrepreneurial lives on an idea.
Migration.
Risks.
Identity.
Loneliness.
Ambition.
Burnout.
Adaptation.
That deserves more honest conversations.
I’d genuinely like to start sharing more of these stories publicly:
the wins, failures, pivots, mistakes, lessons.
Not startup theatre Success is only we live for in American style. We do built something different here in Europe. Based on the same values, we do build different economy. That will have a place for a real human life in it. Safe and comfort.
Real founder life. For a person in business who can be heard.
So if you know interesting founders who came through the Dutch Startup Visa – tag them.
And if you know someone thinking about applying for a Startup Visa in the Netherlands and wanting to avoid mistakes that can easily cost tens of thousands of euros at early stages and millions later, lost equity, years of wrong positioning…
tag them too.
Sometimes one short conversation changes the whole trajectory.
I do 1-hour sessions around founder mistakes people usually discover too late:
https://zcal.co/axlindholm/1hour
#startups #netherlands #founders