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.
The latest NHS FDP stats are out and they show good progress (unless you’re suffering from Palantir Derangement Syndrome).
Trusts using the software have recorded:
▪️ 117,539 additional operations
▪️ 444,373 patients discharged from hospital faster (including a 16.78 per cent reduction in discharge delays for long-stay patients. Remember that debate?)
▪️ 5.68% improvement against the 28-day Faster Cancer Diagnosis Standard
▪️ 1,164,821 patients have been safely removed from waiting lists, prioritising care for those who need it
This is all whilst maintaining the highest standards of data security and privacy, and delivering a cost-benefit ratio is £4.92 for every £1 spent.
"Palantir is a great company," EU defence chief @KubiliusA tells #EuropeToday.
It comes as Lithuania and Poland are reportedly set to partner with the controversial American company, criticised for its use of AI in military operations and mass surveillance.
most of the time, the conversation in Europe isn’t *actually* about safety or data protection
that’s what they want you to believe. it polls well & gets clicks
its *actually* about money. you think the local IT provider cares if the govt system is shit? nah bro. they want to get paid
they don’t want tax dollars (their slice of the pie) flowing to American companies.
NGOs & “experts” are hired guns. social media bots drive engagement and make it feel really real. hide yo kids, hide yo wife tactics
in the end, the biggest losers are the tax payers. nothing gets better. key institutions continue to fail citizens
good luck out there
Angry Twitch streamers are trying to cancel other Pokemon streamers because ModRetro "arms the IDF", "kills Palestinians", etc.
The funny thing is that while I do personally support Israel and the continued existence of a Jewish state, neither Anduril nor ModRetro have actually sold anything to them. Twitch/Amazon, on the other hand...
Nothing screams low-information voter quite like pleading with people to Stand with Palestine and Stop Genocide by subbing to their Twitch, lmao
@luciascarlet The only real hope would be to explicitly exclude mental health coverage from the insurance.
Otherwise its either the blacklist or massive additional costs.
@_Lefyri@luciascarlet You either give them a waiver, so they can look into your history. Otherwise, they'll reject you outright.
If you lie on your application or omit this information, they will ruin your life.
At 250 billion DNS cache entries, one wasted byte costs 250 GB of RAM.
Five Rust optimizations later: 100 TB freed, inserts 43% faster, lookups 19% faster.
We didn't trade speed for space. https://t.co/mMyOYnj0mQ