Robert

@BasedBiohacker If you had just the link in the bio and didn’t source under every post I’d be more inclined to buy from your link. But am in agreement with every post I see of Yours

pinealon is the most underrated peptide for ageing people who want the best possible cognition and recovery, and it's NOT EVEN CLOSE (well, paired with epitalon it's even better) but pinealon on its own: - deepens REM and restorative sleep stages - restores melatonin production to youthful levels - activates telomerase: the enzyme that slows cellular aging at the DNA level - sharpens cognitive speed and working memory - enhances synaptic plasticity and neural connectivity - protects neurons from oxidative damage accumulation - normalizes cortisol and improve stress resilience - reduces brain fog and the afternoon energy collapse it's a bioregulator, and it was made in the 1980s. have been used clinically in russia for decades, but ignored by western pharma because they simply can't patent it. it's simply made of glutamic acid, aspartic acid, and arginine. all naturally occurring enzymes in your body, and the tripeptide can be found in the pineal gland. this isn't medical advice. big pharma, please don't sue me into oblivion.

@benatkin @baltabaev Quickly approaching the Tower of Babel otherwise, where we begin to speak and think no differently than each other, and feel as if we are making incredible bounds of progress

@baltabaev @bigmrrobert Imagine AI training humans too. You might not trust their work as much, but you might trust their work more than that of the AI alone. Gets me thinking about Vernor Vinge novels.

I’ve spent well over 10,000 hours studying math in my life, yet I can’t understand these proofs, at least not without weeks of digging deep into each topic. What’s more, none of my math PhD friends know much about these problems either, and they can’t verify most of them without working directly in the field (yes, math is VERY diverse). LLMs are getting smarter than the experts themselves, and I’m not sure we have enough bright human minds to verify everything that will come out of them in the coming years. Remember when we compared AI intelligence to PhD students? I think we’re past that.

@benatkin @baltabaev I trust AI a lot actually, but there does come a time when we have to decide what training data means

@baltabaev Got it, I thought the tweet you were quote tweeting said LLMs solved proofs that nobody has been able to prove ever

@bigmrrobert There are people who can validate it in reasonable time but these experts are very few. Probably hundreds globally for most of these problems. Random math PhD won’t be able to tell within 1-2 hours because they work in a different domain.

@Dan60784096 @baltabaev Check the tweet he is quote tweeting though

@bigmrrobert @baltabaev I think he’s just saying that he and his friends haven’t specialized in the domain of the problems that were solved. Maybe these were niche problems barely any people were interested in. Someone who studied the problems from this list probably could validate it.

@Rahll @baltabaev I’m in agreement, but someone please correct me, how do you validate a proof that a large language model released if a human cannot validate it themselves? Even if automated via computational techniques that are outside my domain, how do we determine that as truth?

@baltabaev Lmao, why are we glazing things that are unverifiable?

@claudeai It’s ok to admit you are afraid of Codex and Kimi bro

@steipete I’m headed back to the farm

Are we still talking loops or did we shift to graphs yet?

@bcherny 1k agents doesn’t seem reasonable at all yet at any scale

I talk to engineers at other companies every day and hear the same thing: one person is 10x'ing their output with Claude but the rest of the org hasn't caught up. Watching teams adopt AI, I keep seeing the same 4 steps. I mapped them out here: Steps of AI Adoption claude.ai/code/artifact/…

@dwhitedesign Sounds like legends of Zelda breath of the wild sounds

Most UI sounds still feel cheap because they’re just MP3s. built synthesised tactile feedback that generates on the fly. crisp clicks, satisfying toggles, proper error tones. 2kb. zero files. zero dependencies. cuelume-site.pages.dev

@MaxLenormand Hi can I grab a preview? Looks awesome

We're still building some of this out, it's not fully ready to ship but hit me up and I can send you a preview if you want to test this out!

I made a visual viewer for HTTP range requests when exploring a Cloud Optimized Geotif from a remote server As you pan around, it shows you exactly which tiles are being pulled and exactly where the bytes are stored

@JMetz08 @appltrack Same reason many us are renting RSeries AWS instances for with 1500GB, heavy GIS work, tile pipelines, and letting multiple subagents do these concurrently

@appltrack What would I possibly need 1.5TB of RAM for ?

Apple's M7 Ultra chip coming in 2029 is rumored to support 1.5TB of RAM. This would make the processor much more capable for on-device AI.

@whileitravel @Breedlove22 You say this on every post man

@Breedlove22 Peptides will destroy your stem cells potentially taking decades off your life. You should talk to Dr Jack about this Robert.

Everyone keeps asking about my experience with fat loss peptides. So at 40, here's every one I've tested, with the results and side effects each gave me (A THREAD): 1. Retatrutide

there are a lot of benchmarks that suggest 5.6 sol is the best model in the world right now, but the most reliable way to tell is that elon is obsessed with me again

GPT-5.6-Sol sits right at the center of this.

@joshmanders @tropicarls @Bolmercl @theo I’m sorry can you type it slower for me? I don’t understand

@bigmrrobert @tropicarls @Bolmercl @theo I’m typing this message very slowly because we know you can’t read that fast. He said, why this tweet the one to cause this message and not any others?

The Bun rewrite in Rust would have cost about $165k in tokens if it was done over API

@tropicarls @Bolmercl @theo What’s the question?

@bigmrrobert @Bolmercl @theo His question is right there…

@TasmapApp I believe in you

@bigmrrobert That's pretty hard TBH.

✨ Data Import is now available! Tasmap now supports CSV, TSV, XLSX, GeoJSON, GPX, KML, and KMZ data formats. You can now migrate your maps from Google My Maps or most major map platforms with just a single click.

@Bolmercl @theo What’s your question?

@bigmrrobert @theo this one why?