Jonathan Melhuish

@SokobanHero @Teslarati Yes exactly. Seems like people still aren't getting their heads around how AI works. "Tesla FSD has learned to..." would be much more accurate!

@Teslarati It’s not like a specially implemented feature; it’s a behavior learned from the training data.

Tesla appears to be implementing some sort of feature that will now pull over if someone is tailgating you to let the car by Really cool feature, definitely get a lot of this from those who think they drive race cars

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow. Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes. As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now. It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

@DevinOlsenn Oh you think *that* is farting around? [enables fart mode]

I think you can pinpoint the moment that she becomes a lifelong Tesla FSD fan.

@jamesdouma Only one cockpit?! 😲

grok imagine is pretty fun

Airliner designed by ADAS "experts".

@keodark @bcherny If you launch claude --chrome it can control the browser from the terminal version of Claude Code. Would that work for what you need?

@bcherny I want to do this kind of in-browser testing with my app (currently uses Playwright MCP, which isn't the same), but the Claude Chrome plugin can't "talk to" Claude Code running in terminal. I could set up a janky message bus, but why can't Claude Code teleport into the plugin?

I'm Boris and I created Claude Code. Lots of people have asked how I use Claude Code, so I wanted to show off my setup a bit. My setup might be surprisingly vanilla! Claude Code works great out of the box, so I personally don't customize it much. There is no one correct way to use Claude Code: we intentionally build it in a way that you can use it, customize it, and hack it however you like. Each person on the Claude Code team uses it very differently. So, here goes.

@bcherny Ahh, looks like it only works if you run Claude Code in a "proper" terminal, totally outside vscode. Had to find that out via trial and error, sadly Claude had no idea.

@bcherny Is there any way I can allow Claude Code to use a browser to test the code, like you do? This is the only thing that Cline does that I've not figured out how to replicate in Claude Code.

@DoublepPradhan @wholemars @Tesla_AI Exactly. It's weird how Americans are so worried about stopping.

@wholemars @Tesla_AI It does not matter even if someone is behind you. they should stop on time and never tailgate. but the problem is, humans tailgate, react late mostly because distracted just a bit All cars Tesla FSD = no accidents..that is the goal

FSD edge case from my Dad for @Tesla_AI “I was going 80 mph on FSD and a polythene sheet was airborne and got stuck to my car. FSD mistook it for a solid object and braked really hard. From 80 mph to 24 mph. Luckily the car behind me was far away”

@ModdedQuad Nolandlink

Book update: it’s happening. I’m working with a writer to help tell my story, and while we’re still early… I’m already stuck on the title. Neuralnaut 2.0? Eve & I? A Modded Quad Turned Cyborg? Please help before this turns into a 47-option spreadsheet. 😅📚🤖

@DevinOlsenn LLMs are fundamentally bullshit generators

Maybe my Grok is broken, but this definitely didn’t work for me… and then things got weird 😂

@bryan_johnson If you're going to ghostwrite some of the posts, you should probably at least agree on whether sentences start with a capital letter 😂

note to commenters: i’ve been at this for six years now and the tone, texture, and type of commentary is predictable. many people have a strong, knee-jerk reaction of hate because they feel left out, unseen, or down and out. i empathize with that, and i don’t judge you at all. in time, the initial anger quiets, the ideas are digested, and the overton window shifts. if you’re angry, point that anger at the companies printing money while engineering your addiction to their products. they're death machines hurting all of us. reclaim your agency. don't eat fast food or junk food. don't watch porn. put limits on your social media. don't gamble, drink alcohol, smoke, or vape. refuse to do anything that causes you to lose your power. i am your friend, and i want what's best for you. i will continue to freely share everything i learn and stand in your corner.

I’m going to try and achieve immortality by 2039. One year of time passes and I remain the same biological age. I invite you to join me. The search for the fountain of youth is the oldest story ever told. It’s been the dream of dreamers for millennia but always painfully out of reach. For the first time in the history of life on earth, in just the past 24 months, the window has opened for a conscious being to realistically strive for this goal. It is an absolutely insane moment. We currently do not know how 2039 immortality will be achieved. There are new, promising therapies that can turn back the clock decades, but they’re buggy. Sometimes they mistakenly cause cancer. We gotta fix that. But we know immortality is possible because nature has already solved it. This isn’t a physics problem like trying to travel faster than the speed of light, it’s a biological engineering problem that evolution has cracked multiple times. The freshwater hydra constantly regenerates its own cells and doesn’t succumb to senescence. It is effectively ageless. The "immortal jellyfish" (Turritopsis dohrnii) can revert its cells back to a youthful state and restart its life cycle indefinitely. Lobsters produce an especially active form of an enzyme called telomerase that preserves their telomeres and keeps their DNA from degrading as they age. We need to port the software to humans. 2039 is a reasonable target because of the accelerated, AI-driven rate of innovation. AI is morphing from assistant to scientist.  It is powering current researchers with previously unimagined capabilities to enhance discovery and development. That, coupled with enhanced biomarker measurement, creates a closed-loop system of improvement that will speed things up dramatically. This is what I’ve been doing for six years. As crude as longevity technology is today, the improvements I’ve personally seen are stunning. I started as a worn-down, inflamed, aged 42 year old who’d broken himself on the rocks of American food slop and entrepreneurship martyrdom. Six years later, my body largely operates at elite 18 year old levels. That includes my cardiovascular system, fertility, strength, and hormones. Not all is well though. I have mild to moderate hearing loss in my left ear that we can’t fix and my brain is anatomically age 42 (I’m 48). Still, unreal results. My team and I did this by following the scientific method. We measured the biological age of every organ. We referenced the best scientific evidence on how to slow aging and rejuvenate.  We methodically completed these protocols and measured again. Rinse and repeat. We could do so much more and faster if we had better measurement and better therapies. Both of which are cooking all over the world right now. To speed things up now, I’m currently having thousands of Bryan Johnson organ clones built in a dish. This will allow me to test drugs and other molecules against my biology to accelerate learning and save my body from potential mishaps. Yes, we’ll make mistakes. Hopefully they won’t be fatal. And of course there’s always bad luck to contend with. But I trust in fate and I believe that destiny is going to grant the human race the pleasure of achieving the ultimate: immortality. While immortality would certainly be cool, the real (secret) objective is to focus our collective attention on a positive goal. Something that helps us feel hope and excitement for the future. I personally love the idea of having a child-like mind, 18-year-old physical vibrancy, and a lifetime of wisdom. This 2039 goal is as much about AI as it is about us humans. It’s about how we survive giving birth to superintelligence. It’s kind of a big deal. We haven’t done it before. A lot can go wrong. I figure that one of the best ways to improve the probability that we build safe AI, and don’t kill each other in the meantime, is to transform our shared aspirations from yolo to don’t die. Right now, we are a suicidal species. We do all kinds of really primitive shit. We unnecessarily kill ourselves with what we eat and how we live our lives. Companies make profits from killing other people with their products. We trash the only home we have. We celebrate these things as virtue.  It’s really fucked up and backward. Soon enough we’ll realize just how infantile we are right now. The 2039 goal points us in the right direction. To say yes to life and no to death. Defiance even. If you’re interested in doing this with me, I’ll continue to share everything I do for free. I’m also going to build this out in Blueprint. We’ll help you do exactly what I’m doing, at a fraction of the cost and effort, alongside an aligned and motivated community. I think this is the coolest goal imaginable. I find it hard to believe that of all the people who’ve lived, it’s us who get the opportunity to have this moment. I pray that we have the courage to appreciate the sacredness of our existence. I pray that we will be brave enough to defend her amidst the onslaught of all the forces that would try to end her. We don’t know, but we may be the only intelligent life to exist in our corner of the galaxy. I pray we will be warriors, caretakers and stewards of existence and honor the gift that has been bestowed on us.

@seppi111 @teslaeurope Wow, there were quite a few challenging interactions with other vehicles in that short stretch but it judged them all perfectly. Very impressive that it's already working so well on European roads!

FSD Demo Drive in Hamburg today. So smooth. So safe. I hope it will be approved in Europe soon. Congrats @teslaeurope

@battleangelviv And *so* much less carbon pollution & climate impact

@bryan_johnson That's a film I'd love to watch

I want to hear this story… “A tech mogul obsessed with defeating death invites a young woman into his strange, seductive world of longevity rituals, experimental intimacy, and spiritual biotech. As the night unfolds, his “primary partner” and estranged son expose the fractures beneath the transcendence. Blueprint is a psychological drama about the human cost of immortality, the seductions of belief, and the loneliness that technology can’t cure.”

@lporiginalg Breaking news, humans can love more than one person at a time 😱

@jskolte @bryan_johnson You don't like fairy land? 🧚‍♀️

@bryan_johnson Now the interesting part of how long those effects pas or what cadence or dose would be needed to keep it maintained and not going to fairy land each time 🙃

Three exciting longevity results came back from my first dose of mushrooms: >35% decrease in inflammation, a deeply relaxed state and neuroprotective effects. Inflammation (hsCRP) My systematic inflammation levels dropped from elite to undetectable. This decrease is significant, aligning with clinical reports of reduced inflammation markers (CRP, TNF-alpha, and IL6) seven days post high dose psilocybin. + pre-mushrooms: 0.23mg/dL + post-mushrooms: < detection limit 0.15mg/dL + >35% decrease Deep low stress state (HPA Axis) My cortisol and DHEA-S dropped by 42% and 45%, respectively, suggesting a profound inhibition of the Hypothalamus-Pituitary-Adrenal (HPA) axis, which corresponds with the "after-glow" effect. + DHEA-S: 287µg/dL to 157 µg/dL + morning cortisol: 10.3 µg/dL to 5.9 µg/dL While cortisol spikes acutely during the trip, the post-session drop indicates a shift from sympathetic to parasympathetic dominance. This aligns with my sustained relaxed and joyful state. Neuroprotection (Estradiol) Estradiol increased 3x (11.3 to 36 pg/ml). This may be due to peripheral activation of 5HT2A receptors stimulating aromatase activity, the enzyme that converts testosterone to estradiol. A study has shown 5HT2A stimulation in human placental cells can drive aromatase gene expression and activity. My testosterone levels remained unchanged, indicating my body compensated for the minor conversion to estradiol. The increase is within the normal male range and is generally positive, as estradiol provides anti-inflammatory, neuroprotective and cardiovascular protective effects. Conclusion Taken together, these data suggest that a single high dose psilocybin experience can transiently push my biology into a low inflammation, low stress configuration that is theoretically favorable for longevity. Measurement window: 4 days before vs. 5 days post-dose

@erhanerdogan @bryan_johnson Yeah he should probably remind people that when your "study" has n=1 the results should be treated as anecdotal. Anecdotes aren't worthless but they're also not 'proof' of anything!

love following your journey. but the way the numbers are framed in such experiments feels a bit off. these markers swing a lot on their own i.e. CRP can drop 35%+ just from normal recovery in a week. so it doesn’t really make sense to treat such changes as direct outcomes. great storytelling though.

@CyberHooves Your wild art has gone from 2D from 3D. Impressive :)

Roughing in the head fans, 40mm at the front to defog the visor, and a 92mm at the back to defog my head

@yunta_tsai I wish there was some external indication that the car was on FSD; I'd be a lot more relaxed as a cyclist if I could be sure the "driver" was actually paying attention

As a cyclist, I strongly preferred riding near the autonomous vehicles. Full Self-Driving spotted you early, slowing down properly and giving enough space while overtaking gracefully — especially riding along with my family.

@cybertruck 2 years and still not available in Europe 😪

2 years of Cybertruck Thank you to all of my owners (& admirers) – thank you for all the joy, hard work & wild moments so far