I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem.
As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)!
I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work.
It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results?
88ms => 1.5ms
150K allocs => ~500 allocs
Incredible right? Nope.
My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path.
This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput.
The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity.
Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.
@vrexec@KnowledgeUpOnly It will, the moment you are able to sell those shares. This deferral is still a proposal by the way. You might end up paying unrealized gains tax for frozen shares still.
I think not many people is aware of the insanity of exit taxes in Europe. You think you are free? Of course not. Let me share more about NL exit tax.
The Netherlands has an “exit tax” for entrepreneurs.
But calling it a tax is misleading.
It’s closer to a financial exit visa.
Here’s how it works in practice:
Imagine you spent 5 years building a company in Amsterdam.
You took the risk.
You worked nights and weekends.
You paid corporate taxes, payroll taxes, VAT, income tax.
Eventually your consulting/AI firm reaches:
€5M annual revenue
€1.2M profit
€3M cash in the bank
You own the shares through a Dutch holding BV.
Now imagine you decide to leave the country.
Not sell the company.
Not IPO.
Not cash out.
Just leave.
Maybe you want:
lower taxes
better weather
geopolitical diversification
to raise your children elsewhere
or simply freedom of movement
The moment you emigrate, the Dutch state pretends you SOLD your company.
Even though you didn’t.
This is called the “conserverende aanslag.”
A fictional sale.
On fictional gains.
Generating a very real tax assessment.
And the numbers get absurd very quickly.
Example:
Your company:
generates €1M+ yearly cash flow
has €3M sitting in the bank
gets valued at ~€6M by the tax authority
The Netherlands then calculates:
“Okay, your shares increased by €6M.
You owe Box 2 tax on that.”
Current rates:
24.5%
then 31%
Result?
You can receive a tax assessment approaching €2 MILLION…
…despite not selling a single share.
Read that again carefully.
No liquidity event.
No acquisition.
No money in your pocket.
You simply changed residency.
Defenders of the system immediately say:
“Yes, but payment is deferred.”
This is where things become psychologically fascinating.
Because the state effectively says:
“We won’t collect immediately.
But we now have a permanent claim attached to your company.”
A kind of invisible mortgage on your future.
And it can remain there indefinitely.
Then come the triggers.
If later:
you distribute dividends
you restructure
you sell shares
you liquidate
you violate compliance conditions
…the Dutch tax authority can activate collection.
Meaning:
even after you leave,
the Dutch state still shadows the future of your company.
Now let’s talk about the truly dystopian part:
GOODWILL.
For founder-led consulting firms, the value often IS the founder.
The relationships.
The expertise.
The reputation.
The intellectual network.
The ability to close clients.
In many cases, if the founder disappears, much of the business disappears too.
Yet the tax authority can still argue:
“No, no. The company itself has transferable enterprise value.”
So now you enter a surreal negotiation where bureaucrats attempt to price:
your future reputation
your future relationships
your future earning power
your future client trust
The state literally tries to quantify your human capital before allowing you to leave.
And this is not fringe theory.
Dutch valuation teams use:
DCF models
EBITDA multiples
goodwill assumptions
future cash flow projections
…to estimate what your company might be worth.
Not what you sold it for.
What they THINK it is worth.
Imagine explaining this to an American entrepreneur:
“You can leave the country, but first we calculate a hypothetical sale of your startup and issue a multimillion-euro tax claim based on unrealized gains.”
They would think you are describing the Soviet Union.
And before people scream:
“but all countries tax!”
No.
This is different.
This is taxation attached to MOVEMENT.
That distinction matters.
A normal tax says:
“You earned money.”
An exit tax says:
“You are attempting to leave.”
That changes the philosophical nature of the relationship between citizen and state.
At some point it stops resembling taxation and starts resembling permissioned mobility.
Especially because the system creates behavioral lock-in.
Founders begin asking:
Should I leave before growth accelerates?
Should I suppress valuation?
Should I avoid accumulating cash?
Should I avoid long-term contracts?
Should I move BEFORE success?
Think about how insane that is.
The tax system starts influencing:
where founders live
how companies are structured
when growth is pursued
whether capital stays in the country
And Europe wonders why founders eventually relocate to:
Dubai
Singapore
Switzerland
Miami
Cyprus
The real irony?
Politicians claim they want:
innovation
entrepreneurship
AI leadership
startup ecosystems
But the moment a founder becomes globally mobile and valuable, the state says:
“Not so fast.”
Build in Europe.
Scale globally.
But if you leave, we’ll first calculate your hypothetical future and attach a tax claim to it.
That is not a free market. We are not free.
Is it confirmed that she is actually working with Sandfall Interactive? In the photo, it looks like they are sitting somewhere and she just walked in and snapped a shot. She might be attempting to tie herself to them.
It would be disastrous if Sandfall actually allows her to have influence over any of their work. Anything she touch turns rotten. I've been boycotting her for a very, very long time.
@GergelyOrosz I do the same. Really sorry for the false positives I hit. I just can't stand wasting my time reading some ai slop someone not even bothered to actually write.
Your post is the false news. It's not cancelled. It will be approved at the upper house after some possible amendments. These amendments are expected to be some stuff like "if you have private shares you can't sell, then those can be deferred" etc.
Current cabinet will **start** a discussion to change the unrealized box 3 tax to a realized one (sucks less, but still sucks considering we already pay %49,5 top income tax rate).
Nothing is proposed yet to change this incoming unrealized box 3 taxation.
@TrailsideTrades@synthwavedd I think you meant to reply to OP. I'm talking about GitHub Copilot here. It's not a speculation that Opus models are removed for Pro users. This happened last week. You can see the details here: https://t.co/FSzHdMTzGw
@SaladBarFan You should read "From Third World To First". It's a long read but shows the deep level of challenges Singapore faced and what LKY and his friends did to overcome them. It's seriously impressive what they achieved.