Anton Becker

Geldentwertung, Inflation, Kaufkraftverlust

@levelsio There should be a moratorium though until they can try again. A few years at least.

HOORAY!!! 🥳 They will keep trying, we'll keep fighting to reject Chat Control

@AlSultan_Meriam @daviddorg Anything is portable if you're strong enough

@daviddorg Is it portable? Doesn’t seem like it.

This is my “Screenless phone” I will be using for my year without screens I write on paper I scan it in It follows my commands: send a text, send an email, print me this beautiful artwork (I love John William Waterhouse) The future is near

@MicrobiomDigest @drkeithsiau Once heard about a family physician who suspected jaundice in his first few patients of the day. Then realized it was the new yellow curtains.

@drkeithsiau As a child, my parents were very concerned about my hands turning blue. It was just my new blue jeans 🤣

In BBC news today, a 42M from Derbyshire, UK, was rushed to hospital after waking up looking blue, “like a character from Avatar”. A friend convinced him to rush to hospital where concerned ED staff put him on oxygen. All was reassured when a medic rubbed his arm with a swab that turned blue, revealing that the suspected cyanosis was blue dye from his bed sheets.

@aleabitoreddit @RB1Patrick At which point would you cut your losses?

@RB1Patrick Yes $RDDT is one of my core longs around $137. I’m just upset it went below my cost average.

$RDDT is getting ridiculous. Looks completely mispriced. Now down 40% over the past few months. If you strip out carry-forward losses, their net profit is ~28% of revenue, which is absolutely enormous. And they’re growing forward revenues 50%+ Y/Y after 70%+ Y/Y growth. If you ever look at $META, you know how much revenue can be optimized/user. There’s an incredibly high ceiling for monetization with Reddit. It’s already derisked since IPO since Reddit is now one of the fastest growing and highest margin companies in the market. One day if it pulls a $CRCL post earnings, we’ll look back and wonder how this was valued at $24 billion MC.

@samcoward @levelsio Traffic lights are a mere recommendation in NYC

@levelsio You should see New York. It’s bananas here. Seems like there’s only 1 rule, don’t hit anything. Nothing else is sacred. Hundreds of illegal turns a day just outside my window. traffi.skcfi.dev

Now that I got my driver's license Seeing people drive like complete goons every day now I think our #1 priority should be to get every car brand to self driving ASAP on a worldwide gov mandate Humans are way too retarded to drive

"How long until we cure aging" "Like two years" "How long until you can fix my shoulder" "Oh heavens no. That's impossible"

@Tounsi_CHF @chribreuer 🎯

That’s too absolute. Switzerland did produce important intellectual and artistic figures:Rousseau, Pestalozzi and Burckhardt, Jung, Dürrenmatt pr Giacometti and so on. So "no poets, no philosophers, no composers" is just false. But there is still a reason why Swiss high culture feels thinner or less visible than German or Austrian high culture. Switzerland never had the same kind of large and centralized cultural machine, no imperial court, no Vienna, no Berlin, no Paris-like capital concentrating prestige, patronage, academies, and symbolic power. Its elites were mostly local, urban, fragmented, bourgeois, and pragmatic rather than courtly and civilzational in ambition.

One interesting thing about the Swiss is that, although they’re essentially “super Germans” in terms of efficiency and economy, they never developed the same kind of high culture as Germans or Austrians. No composers, no poets, and no philosophers

@lulumeservey @sukh_saroy Study about GPT, written by GPT, tweet-promoted by GPT. We have come full circle.

@sukh_saroy The irony of seeing this study and using Claude to write you a tweet about it

🚨BREAKING: If you've used ChatGPT for writing or brainstorming in the last 6 months, your creative ability may already be permanently damaged. A controlled experiment just proved the effect doesn't reverse when you stop using it. 3,302 creative ideas. 61 people. 30 days of tracking. Researchers split students into two groups. Half used ChatGPT for creative tasks. Half worked alone. For five days, the ChatGPT group outperformed on every metric. Higher scores. More ideas. Better output. AI was making them better. Then day 7. ChatGPT removed. Every creativity gain vanished overnight. Crashed to baseline. Zero lasting improvement. But that's not the bad part. ChatGPT users' ideas became increasingly identical to each other over time. Same content. Same structure. Same phrasing. The researchers called it homogenization. Everyone using ChatGPT started producing the same ideas wearing different clothes. When ChatGPT was removed, the creativity boost disappeared -- but the homogenization stayed. 30 days later, same result. Their creative range had been permanently compressed. Five days of use. Permanent damage 30 days later. A separate trial confirmed it. 120 students. 45-day surprise test. ChatGPT users scored 57.5%. Traditional learners scored 68.5%. AI reduces cognitive effort. Less effort means weaker encoding. Weaker encoding means less creative raw material. You're not renting a productivity boost. You're financing it with your originality. The interest rate is permanent.

@WesRoth Yeah also, he's completely wrong

Remember when everyone said AI was going to completely replace radiologists? Dario Amodei just broke down exactly why that hasn't happened yet and what it means for the future of your career. Instead of wiping out the job, AI just ate the highly technical part (reading the scans), forcing humans to pivot entirely to the "human touch" (patient care). Amodei believes we need to adapt to AI step-by-step as it evolves. However, he didn't sugarcoat the endgame: he admitted it is highly likely that AI will eventually beat us at everything, including physical robotics.

@airkatakana It gets even harder if you cross-link some of these articles and create a web of lies. I did this with an invented word and 1 reference many years ago, and it stuck around forever and eventually ended up in some wikis outside of Wikipedia. I think it still lives on to this day.

i’ve spent the past few months generating fake “extinct animals” with nano banana and making wikipedia articles for them. all backed up by book references that don’t exist. not a single article has been deleted yet out of hundreds. the future is here

@levelsio @paulsaladinomd @foundmyfitness Baffles me how well your technique of dry spooning works

@paulsaladinomd 5g is like last year with all due respect 10g is the new standard, see @foundmyfitness's advice, 5g for muscle, 5g for brain 20g if you're traveling for sleep deprivation

Creatine optimization protocol: 1) 5g creatine monohydrate daily. 2) Dissolve FULLY in water (warm helps). 3) Use after workout (some data suggests slight benefit) 4) Consume with carbohydrates (50-100g) & salt (200-500mg) to improve absorption. 5) Use a 3rd party tested brand.

@eugenezelenkov @levelsio You can just set the owner of the env file to whichever user runs the final application (e.g. www-data), and chmod to 600, then Claude won't be able to read it.

@levelsio It is really annoying to confirm every step of it BUT i've seen (and stopped) how claude tried to read my .env file when i asked never do that and now i am a bit scary to do smth like that 🥲

This shortcut has helped me ship way way way faster, I've been using it manually before but now I just run every Claude Code bypassing permissions I was getting sick of answering endless questions and giving permissions, just do the job already I'm 100x faster now echo 'alias cc="IS_SANDBOX=1 claude --dangerously-skip-permissions"' >> ~/.bashrc source ~/.bashrc

@ouranoscapital @yudapearl Was joking. Was actually my first thought as well

@dr_becker @yudapearl I'm actually a CS major btw

There is an error in this image -- I've never said "The mind is not a Turing machine". I strongly believe the mind is a machine; surely not a moving tape, but a machine nevertheless.

@phoenixnflames @Papa_Heme @EvidenceOpen ✨VPN✨ (also free)

@Papa_Heme @EvidenceOpen But, third world issues -

It pains me to say this but OpenEvidence AI is superior to UpToDate (and it’s free!) @EvidenceOpen

@nirbheek @webdevMason Would be my guess too

@webdevMason If you do this too much, you probably get autoimmune diseases or cancer right?

I would LOVE to rid my family of both minor and serious infectious illness, but if the immune system already has a "just do what you do, but better" mode, the obvious question is why it hasn't been adaptive for that to be constantly activated by default

@ouranoscapital @yudapearl Found the philosophy major 😜

@yudapearl What do you mean by machine?

@levelsio Hahaha the whole lower half of Manhattan is a 'Bad Area'

✨ Made a new robot for my site Hoodmaps.com that uses its data to find good and bad areas and mark them It collects: - crowdsourced drawing (like an area is tourist) - crowdsourced tags data like "Pickpockets everywhere" - positive places data like lots of trees, parks, playgrounds, cafes to work, coworkings - negative places data like lots of pawn, money transfer, payday loan shops, homeless shelters or strip clubs And then lets @xAI Grok 4.1 Fast use that data to figure out the best and worst areas of a city I just tested it and I think it works! 😊👌

@AnilMakam I think you overestimate the resolution of current whole body MRI protocols

Another reason why whole body MRIs will lead to net harm

@levelsio Cerebras is pretty damn fast, but they only have open models. No Claude unfortunately

I could work with 1 single Claude Code tab if it responded 100x as fast But now it takes 1-3 minutes sometimes Once that's 1-3 seconds it's fast enough for me to just keep going in a single tab But now I get blocked by how relatively slow AI is (I mean way faster than a human but you get me)