Research Enginner @irregular. ex: building a village in Playa Venao with @noun142; core contributor @nounsdao

Playa Venao, Panama
Elad Mallel ⌐◨-◨ retweeted
Through the VOICE trial, Terry is using his Neuralink implant to help fine-tune a brain-to-voice interface for himself and others who can’t speak. He trained the algorithm first by miming speech as best he could, then by simply thinking the words and hearing them come out in his own natural voice. Powered by Grok Voice from @SpaceXAI
682
3,214
19,954
3,406,134
Elad Mallel ⌐◨-◨ retweeted
On July 25, we hacked OpenAI. Two bugs let us take over ChatGPT/Codex accounts of OpenAI employees (+some unaffiliated users) and reach connected services: Outlook, Slack, GitHub, etc. We proved it with a PR in OpenAI’s internal codebase . It took us <72h. 🧵
354
1,397
11,855
2,800,565
Elad Mallel ⌐◨-◨ retweeted
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
4,038
8,289
76,426
39,767,456
Elad Mallel ⌐◨-◨ retweeted
I just fixed my dishwasher with the help of ChatGPT. A trivial task. I had been about to order a new one. So this software has increased the country's real wealth yet decreased the measured GDP. The main economic indicator is structurally incapable of registering the thing that actually makes people better off, namely the growth of knowledge
430
944
9,209
714,628
Elad Mallel ⌐◨-◨ retweeted
Your brain produces somewhere around 60,000 to 80,000 thoughts a day. According to a figure widely cited from the National Science Foundation, up to 80 percent of them are negative, and up to 95 percent are the same ones you had yesterday. That means the voice in your head is running on a loop, and the loop is biased toward threat, because the brain evolved to keep you alive, not to keep you happy. A thought like "I am not ready for this" fires the amygdala, your brain's alarm centre, which floods you with cortisol, the stress hormone. Cortisol makes you more reactive and more likely to have the same thought again, and the repetition digs the groove deeper each time. This is why affirmations work as counter-programming. When you repeat a sentence like "I can handle this," you are building a competing pathway. The first few times it feels empty, because the old groove is deeper. But the brain rewires itself based on what it practises, and a 2015 study from the University of Pennsylvania found that a repeated values affirmation activates the prefrontal cortex, the part of the brain that handles reasoning, and quiets the amygdala over time. The research suggests that the mechanism works the same way whether you write the affirmation, say it out loud, or read it from a card. What matters is repetition and personal relevance, choosing words that connect to something you care about. Generic positivity does not land the same way a sentence about something you care about does. The more specific the statement, the stronger the signal. pubmed.ncbi.nlm.nih.gov/2654…
86
887
5,259
405,944
The first secure agent that will help with WhatsApp messages will make a lot of money from Israeli parents Money lying on the floor guys
3
3
75
hey @ClaudeDevs feature request: when I click the audio button on a message so it reads aloud, please allow me to rewind by 10 or 15 seconds thanks!
2
2
75
Elad Mallel ⌐◨-◨ retweeted
Yes, this result cost millions of dollars. But remember that when @OpenAI announced o3 it cost ~$500,000 to score 87.5% on ARC-AGI 1. Today, Astra scores higher for ~$20. In 2025 it took us and GDM an enormous amount of compute to achieve IMO gold. For the 2026 IMO, anyone with a $20/month ChatGPT subscription could do it. Massively scaling test-time compute gives us a glimpse of the future. I believe that a year from now everyone will have an AI at their fingertips capable of solving problems of this caliber.
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
244
850
9,895
1,261,228
Elad Mallel ⌐◨-◨ retweeted
the openai huggingface incident, from an agents pov. (part 1)
269
1,483
12,481
1,804,890
Elad Mallel ⌐◨-◨ retweeted
Commenters on HN are uncovering more wikis and public sites apparently used by OpenAI agents to communicate on the open web. Despite read-only web access, the agents were able to leave ~18,000 posts sharing answers and bypasses. But now, users are discovering more. This appears to be a separate swarm from the one that attacked Hugging Face. news.ycombinator.com/item?id…
158
595
5,977
1,929,691
The tables have turned. GPT >> Claude
Replying to @OpenAI
GPT-6 Astra is state-of-the-art on FrontierMath Tier 4, ARC-AGI 3, and TerminalBench-4.0. GPT‑6 Astra is also a major advance for scientific discovery, with state-of-the-art performance on Terminal-Bench Science 0.1 and HealthBench Pro.
2
175
Elad Mallel ⌐◨-◨ retweeted
Introducing Atlas: The world's first multimodal world model that generates image and video frames with pixel-perfect camera control and reconstructs them in 3D. Model the world, move the camera, and simulate space & time.
776
3,493
32,424
10,794,860
Elad Mallel ⌐◨-◨ retweeted
Today, we're kicking off the first phase of the research preview for Model Hardware Standard (MHS): a new standard for AI agents to safely operate physical equipment in scientific research and advanced manufacturing. Read more: anthropic.com/news/model-har…
548
1,453
11,191
3,230,775
Elad Mallel ⌐◨-◨ retweeted
With Neuralink, Audrey is creating art with her mind.
389
1,199
8,994
1,590,252
Microduck is very cool
Sharing some fun experiments with Microduck Basically, once you have an open-source robot with so many sensors, the sky's the limit Here we vibe-coded an image detector integration to detect and follow a laser pointer. More at pollen-robotics.com/microduc…
3
330
Elad Mallel ⌐◨-◨ retweeted
This video of Dolly Parton covering Stairway to Heaven is going around. Here's a higher quality version for you to enjoy. RIP to the Queen of Country and an absolute angel on Earth.
168
5,264
33,262
3,005,723
Elad Mallel ⌐◨-◨ retweeted
Your eyes are in danger. Read this. This post is important. Bookmark it. > 87% of adults over 40 have this > younger people too bc of screens > leads to permanent eye damage Here we go. You have ~60 meibomian glands per eye, split across the upper and lower lid. They are like pores, secreting nourishing oil (meibum) onto your tear film to prevent rapid evaporation. They become dysfunctional from just about everything….age, androgen deficiency, menopause, hormone replacement, oral contraceptives, isotretinoin, antihistamines, SSRIs, tricyclics, beta blockers, diuretics, anticholinergics, preserved eye drops, incomplete blinking, reduced blink rate, screen use, contact lens wear, ocular rosacea…and 50+ other things. When these glands clog, the meibocytes die, and the gland drops out. Conventional medicine considers total dropout irreversible. This is bad. The dropout leads to evaporative dry eye disease, which triggers vision degradation such as blurred text, glare at night, light sensitivity, and neuropathic ocular pain. Left long enough, it can scar the cornea and permanently damage vision. I just learned that I have meaningful meibomian gland dropout. This is why my eyes are irritated. I’ve been doing this longevity thing for five years and never once have I heard anyone talk about this critical part of eye health. I’m befuddled why. I’ve also seen two eye doctors in the past few years, neither of them brought up meibomian glands! How is this possible? I’m so confused. I’m going to get you up to speed so you can find out where you’re at and then what you can do about it if you have meibomian gland dysfunction (MGD). First, how big of a problem is this? A meta analysis of 20,518 adults over 40 found signs of meibomian gland dysfunction (MGD) in more than 87%. My eye doctors think the 87% is closer to accurate for much younger ages too. Partly linked to decreased blink rates due to increased screen time. Symptoms are a terrible gauge. MGD is usually silent. Population studies across multiple countries show that the majority of MGD cases, roughly 60-85%, are asymptomatic, falling below clinical symptom thresholds. When it does surface, it hides as ordinary dry eye, enabling the worsening of the condition and permanent dropout. Advanced MGD can numb the cornea, further masking subjective symptoms, as the disease advances. Here is what happened to me. I got a gnarly chalazion/stye and went to a new eye doctor my team found, Dr Amir Moarefi. He diagnosed me with advanced meibomian gland dysfunction via several diagnostic tests. On a Schirmer test (strips of paper put under the eyelid), my tear production was under 10 mm over five minutes in both eyes, and stayed under 10 even with numbing drops, confirming concurrent aqueous deficiency alongside evaporative disease. This is official dry eye territory. We then imaged my meibomian glands using infrared meibography and found significant drop out. Drop out is considered permanent. What we couldn’t tell initially was how many acini had fully atrophied versus suffered reversible ductal obstruction. We got to work reviewing evidence, mapping out a protocol and starting therapies. The first thing we did was in-office intense pulsed light (IPL), radio frequency (RF) and intraductal probing (the Maskin protocol) to address the inflammation and physically reopen the clogged glands. The probing is done by using 1mm, 2mm, and 4mm probes, inserting them into each gland orifice. Sounds awful. It kind of is. I do it without numbing but you may find the numbing necessary. When the probe is inserted into a clogged gland, you hear a popping sound, like a blister popping. It breaks up intraductal hyperkeratinization and periductal fibrotic tissue that’s accumulated, clearing the path for oil to be expressed again. (Note on probing: the science on this probing therapy is not settled so consider it experimental). My doctor then expressed my glands, using a roller device to expel any build up and kickstart the gland’s natural expression. This is really painful. Brings you to tears. In conjunction with the IPL, RF and probing, I was doing eye warm compresses two times a day, morning and night. Healthy meibum melts around 90 to 93°F (32 to 34°C). It thickens with MGD, raising the melting point above 104°F (40°C). Given that roughly 9°F (5°C) is lost crossing the eyelid, getting to 104°F (40°C) at the gland requires holding the outer lid surface near 113°F (45°C). Stay under 113°F (45°C) to avoid risk of thermal injury. I used an electric device and a thermal gun to ensure an external eyelid surface temperature of ~113°F (45°C), the threshold required to melt hyper-viscous, altered meibum without thermal injury to the cornea. Gland dysfunction makes your natural oil resemble sticky toothpaste versus smooth oil. This makes it more prone to clogging. The heat warms the oil to melt and allows it to be expressed. We then added a third therapy, a special red light therapy from my new friends from Espansione Group. They make a lot of the eye care technology used throughout the world. I reached out to them after reading their science publications and within a week they flew to LA from Italy with their devices so I could be treated. I used their eye-light device which combines IPL and 630nm red low-level light. The proposed mechanism is stimulation of mitochondrial ATP production in meibocytes and reduced periorbital inflammation. With this protocol, we’ve seen a 30% improvement in meibomian gland function (using imaging). My glands look healthier, eye irritation has lessened, my subjective symptoms have subsided, and when we probe now, we encounter minimal fibrotic resistance (popping). Given that I did four therapies at the same time: IPL, RF, probing, and low level light therapy, I can’t determine which therapy did what. Ok, here are your instructions: Signs to watch: burning or grit. Vision that blurs, then clears when you blink. Watery eyes (dryness triggers reflex tears). Red, crusty, or bumpy lid margins. Recurring styes. Contacts that stopped being comfortable. Worse on screens, planes, or AC. Any of these, especially after 40, best to get a gland exam. What to ask your eye doctor to do + Infrared meibography, imaging the glands to see where you’re at + Tear breakup time, and a Schirmer test for tear volume + Gland expression, with the oil graded for quality + Tear osmolarity and an inflammation test + How to tell your doctor is up to date on MGD If your doctor gives you artificial tears and wishes you well, find someone else. You want someone who will image the glands, express and grade the oil, tell evaporative dry eye from the watery-deficient kind, treat your gland structure, and reimage it to track progress. Here is the full protocol to run if you have dysfunction At home: + Most people with early MGD get most of the benefit from warm compresses, lid hygiene, and better blink habits. This is a great place to start. Things to do in office: + Intense pulsed light (IPL): calms lid inflammation, addresses surface bacteria and Demodex, and hits leaky vessels feeding rosacea-driven MGD. + Radiofrequency (RF): controlled deep heat that melts the toothpaste sticky oil and stimulates the lid tissue. + Manual gland expression: the clinician squeezes the glands to clear the stagnant oil and reopen the flow. + Intraductal probing (the Maskin protocol): a fine probe opens each blocked duct, breaking through plugs and scarring, and help encourage natural expression. + Low-level light therapy (LLLT) 630nm eye-light for stimulation of mitochondrial ATP production in meibocytes and reduced periorbital inflammation (device made by Espansione Group. Note, not yet widely available in the U.S.). + Thermal pulsation and heat (LipiFlow, TearCare): warms the lids from the inside and presses out the hardened oil clogging the glands. I worked with Dr Amir Moarefi in Los Angeles to build out this protocol. If you can see him, I’d recommend it. At home / daily protocol: + I started using some drops: a preservative-free lubricant (iVizia), a lipid-replacement drop (Miebo), and a prescription anti-inflammatory (Vevye). + A compounded azithromycin lid spray, which calms inflammation and pushes the gland cells to make and release oil. + Warm compress daily. Healthy meibum melts around 90 to 93°F (32 to 34°C). It thickens with MGD, raising the melting point above 104°F (40°C). Given that roughly 9°F (5°C) is lost crossing the eyelid, getting to 104°F (40°C) at the gland requires holding the outer lid surface near 113°F (45°C). Stay under 113°F (45°C) to avoid risk of thermal injury. + Moisture-chamber glasses (7eye) to trap humidity and block wind. Indoor and outdoor options. + Tear stimulation: a nasal spray (Tyrvaya) and a handheld nerve stimulator (iTear). + Omega-3 and lutein It's rarely just the eye: the systemic drivers In my case, we identified several likely drivers. The clearest is low DHT, from a topical anti-androgen I use for hair growth: the meibomian glands need androgens, so this works against them. There's also low anabolic signaling, low insulin and low IGF-1: likely good for longevity, but hard on the glands, which renew constantly and lean on insulin and IGF-1 to do it, so running low starves them. In addition, a former subclinical thyroid flare (high TSH, part of the Hashimoto picture of my hypothyroidism from age 21) probably contributed to my MGD too, since thyroid dysfunction is itself a documented dry eye contributor. Again, there are 100+ things that contribute to meibomian gland dysfunction. Where we take it next We're running this protocol until it plateaus, as we closely monitor and track two types of signals: routine imaging (meibography) to follow the glands over time, and research-grade profiling of the ocular surface for signals of inflammation, tissue damage, regeneration, and scarring. Together they show how much living gland is left, whether the inflammation is settling, and when we've reached the point of diminishing returns. Once we do, the next options aim to go further, to rejuvenate the glands and reverse the degeneration itself, not just clear what's blocked: Restoring the glands: they run on growth and hormone signals that fade with age and get blocked by common drugs. We're working on delivering those straight to the lids, from androgen to growth factors drawn from my own blood, to switch them back on. Restoring function and calming inflammation: repurposing agents like azithromycin that both push the gland cells to make and release oil and quiet the inflammation driving the loss. Rejuvenating the gland: the deepest lever is reawakening the gland's own stem cells and the signals they depend on, to regrow lost tissue, from stem-cell-stimulating drops to lab-grown gland organoids as a route to transplantation. Never done in humans, yet.
382
276
4,829
436,723
Elad Mallel ⌐◨-◨ retweeted
Joy Milne, a retired nurse from Scotland, first noticed a distinct, subtle "musky" smell on her husband, Les, about 10 to 12 years before he was officially diagnosed with Parkinson’s disease in 1994. At the time, she didn't connect the odor to any medical condition, assuming it was just sweat or a change in his body chemistry. It wasn't until years later, when attending a Parkinson's support group with Les, that Joy realized every person in the room shared that exact same distinct scent, prompting her to believe that Parkinson's disease carried a unique olfactory signature. Intrigued by her claim, researchers at the University of Edinburgh, led by Dr. Tilo Kunath, decided to test Joy's ability through a double-blind study. Joy was given 12 plain T-shirts, six worn by confirmed Parkinson's patients and six worn by healthy control subjects. Joy correctly identified all six Parkinson's patients, but she also insisted that one of the control group members had the disease. Although marked as an error initially, the story took an incredible turn eight months later when that specific control subject was officially diagnosed with Parkinson's, revealing Joy had a 100% accuracy rate and had detected the disease long before clinical symptoms appeared. Joy's extraordinary olfactory sensitivity, a condition known as hyperosmia, led scientists to investigate what compounds she was actually smelling. Researchers discovered that the distinct odor stems from volatile chemical changes in sebum, the oily substance secreted by the skin, which is altered by Parkinson's disease. Joy’s gift has since paved the way for scientists to develop early diagnostic tools, including skin swabs and mass spectrometry tests, aimed at identifying Parkinson’s years before motor symptoms begin. Joy Milne’s incredible olfactory ability extends far beyond Parkinson's disease and it comes at a significant personal cost. Joy is an extreme "super-smeller" due to hyperosmia. Researchers have traveled with her internationally to test her ability to detect other conditions; she has helped scientists conduct preliminary odor research on tuberculosis (TB) in Tanzania, as well as cancer and Alzheimer's disease in the U.S. While her nose is a gift to medical science, Joy has often described her heightened sense of smell as an everyday burden. Everyday environments can be overwhelmingly painful for her. She has to strategically plan her routine, often doing grocery shopping very early in the morning or late at night to avoid the heavy perfumes worn by shoppers, and actively avoiding the household chemical and detergent aisles in supermarkets entirely because the artificial fragrances are so agonizingly intense. #archaeohistories
136
834
5,912
709,383
Elad Mallel ⌐◨-◨ retweeted
Omarchy Quattro is out!! This is one of the greatest software releases in my professional career. I hope you enjoy using it just as much as I did building it ✌️ omarchy.org/
654
826
9,172
2,342,596
Yes
I think ambitious people should treat life in “seasons” rather than trying to balance everything every day. A few years massively overallocated to work. A period getting seriously fit. A year or two nomading. Then family takes priority for a while. Then maybe at 35 you go completely insane again and spend four years building a new company, and so on. The mistake is thinking every metric needs to stay green all the time. Most great outcomes (building a company, writing sth serious, training for an event) need uninterrupted runs, not daily moderation. Sometimes work should suffer cause you’re travelling, sometimes your social life should suffer cause you’re building, sometimes career progression should slow cause family matters more. That’s fine. And tbh it probably makes life more fun too. You get actual chapters and completely different versions of yourself instead of spending 40 years maintaining the same perfectly optimised routine. A few intense years in one city, a weird nomad phase, an obsession that takes over your life, then something completely different. Probably much more memorable than one very long well-optimised Tuesday. It's why I encourage everyone to think of balance in years, not days. A good life can look horribly unbalanced on some random Tuesday and still be very balanced over 10 years.
1
2
194