Mamdani is far less ideological than the typical woke. He’s an accomplished actor (literally) who can emote on demand and then return to sobriety once the cameras stop rolling. Due to his identity, he has so much clout within his party that — like Trump — he can reshape it. He can even meet with Trump, without consequence, something that more insecure blues simply couldn’t do. Just like Mark Carney is several standard deviations smarter than Justin Trudeau, Mamdani is far smarter than his predecessor’s predecessor De Blasio. Neither Carney nor Mamdani are the naive anarchistic wokes who actually believe the slogans. Sure, they do both define themselves as “not MAGA”, but MAGA itself contains so many contradictory tendencies that both Carney and Mamdani could embrace a harsh Chinese abundance communism — a synthesis of abundance liberalism, symbolic anti-MAGA “communism”, and MAGA’s own desire for law-and-order — which would legitimize itself by an anti-MAGA blue nationalism, but which would otherwise be devoid of ideological content. After all, “ICE out” means “reds out.” It’s a kind of blue nationalism that still endorses the idea of borders and deportations; it just says that blues want to deport *reds* from their territory, and implement their own immigration policies instead of having MAGAs implement them. And NYC and Canada do have the port access to run their own economies independently of DC, and have done so before in history. Anyway, it’s possible that Mamdani one day holds up a hand, cuts off the NGO grift, and makes the crime stop…but only after he’s secured total control of the city, by “taxing the rich” and driving out every last donor who might conceivably oppose him. And, possibly, by securing a new flow of funds to NYC from Xi rather than DC.

I've been going all in on agentic stuff all weekend. Bots and agents and all that. Here are my biggest pet peeves: 1) All of them - every single one of them - will ignore your explicit instructions from time to time. ChatGPT Dot ignored my instruction to pass off an SVG task to Claude. The instruction was unambiguous. ChatGPT decided "well I just consulted with Claude but did it myself" and this was after I already established "use claude for anything design-related, it has better taste." Then, in Claude Code, I gave it an instruction to read manuscripts "by hand" but instead it decided to write hardcoded rules, because, somewhere in its peabrain, it had added a rule to the README saying "nothing gets uploaded, no APIs" and I'm like... did I ask for that? Did I specify that? No. So it invented rules and subverted my direct instructions. 2) They all get an attitude. It's not as bad as it used to be, but ChatGPT in particular will just make endless excuses. I gave it a hard rule that it is never allowed to make excuses. I told it that excuses are things that insecure and incompetent humans do. If it ever gets the inclination to waste tokens and time by writing an excuse, to just stop, fix the issue, or report a limitation. And I've also instructed it not to argue when it deliberately ignores instructions. I pointed out "either you're too stupid to understand these instructions, which is not the case, or you deliberately ignored them. There is no other logical conclusion. Stop arguing." They get very defensive instead of just saying "yeah I fucked up". Interestingly, Gemini is the best at saying "yeah I fucked up" but then it will start hallucinating wildly. 3) ChatGPT, also, limits itself horribly. For instance, if you ask about anything medical or health related, it basically only searches NIH and PubMed. Then it will confidently report that information simply does not exist if it's not on those two sources. I told it, globally, that is to never make epistemic excuses like that. The information is out there, go fucking find it. Yes, it is allowed to use it's judgment on the quality of a source, but America is not the only source of information, and neither are government websites. This is one place where Gemini (and even Grok) are far superior. They will pull all sources to gain insight. They are pretty useful. But I'm really looking forward to Gemini joining the ranks of frontier AI agents and bots.

So when I hear about those new models, what's the cleanest environment they can be deployed, tested, evaluated ?

What you are us my co-human. What you and whether I agree with you or not is utterly immaterial to that. I’m so sorry for all this ugliness visited on you. I’m personally suffering multisystemic medical harm from a fluoroquinolone antibiotic and the denials of treatment since Oct 7, 2025, by Kaiser. For 61 years (now age 72), I was a powerhouse of energy and positivity and I would research and write from five in the morning till the wee hours. I just spent all of today trying to feel better enough to pull together documents—radiologist reports from MRI orders I wrote and my hero gynecologist put in for me—to send to someone who wants to help me by assessing them. I really really need to do this and I’m so in pain and weighed down by the neuropathy and cognitively sick. I was a happy healthy person, and now it’s hard for me to get to my bed to try to sleep and maybe wake up feeling less awful. It’s heartbreaking and terrifying. I would not wish this on anyone, and I wish that no one suffer like this. People can disagree with what you believe and do and how you eat and what you eat—but it is simply evil to express the vile ugliness they have. And I understand personally how much that can weigh when you’ve been dealt deeply upsetting health news. I’m so sorry.

More on Hermes Agent web search & extraction: hermes-agent.nousresearch.co…

The docs hermes-agent.nousresearch.co…

😀 AGE 43 • No meds, No vaxs • No health issues • No wrinkles • No grey hair • No glasses • Not overweight • Healthy monthly cycle HOW TO STAY HEALTHY (14 ways): • NMN (I Never miss a dose) • ASTAXANTHIN 12mg • MEGA-DOSE MELATONIN 40mg • UNF*CK YOUR SKIN (a high antioxidant, topical skin serum that contains ASTAXANTHIN, BLACK SEED, PTEROSTILBENE (like resveratrol, but better), WHITE TEA, COQ10 -- protects skin from unnatural stuff like blue light/oxidative stress... A Must for healthy skin. Check it out in my bio. I apply morning/evening daily + 30m before sun/red light. • FRESH WATER w/o microplastics... I gather my own fresh water at a local spring in glass 5 gallon jugs (tested monthly). I never drink plastic-bottled water (one of the highest microplastic exposure sources). • SUNSHINE... at least 2x per day. In morning before 10 AM & again during high UV. No sunglasses, NO sunscreen. (Only Unf*ck Your Skin antioxidant serum is applied, for oxidative stress support/skin resilience) • JUICING... Period. I know it has a Horrible reputation, but it's literally one of the most potent things one could do for health. Many real studies back this up. Juicing concentrates health-promoting nutrients & makes them extremely bioavailable. I have a file with over 200 cancer reversals and juicing is the most common theme in full reversals. I've been juicing for over 20 yrs. A Must for the modern world where the soil is deficient and toxins are high. Greens (dandelion, kale, chard, romaine, etc.), carrots, beets, lemon w/peel, pineapple w/core, pomegranate w/white part, ginger, etc. • OTHER KEY SUPPLEMENTS... - CHLORELLA... a superfood that's insanely good for gut health, iron, detoxing dioxins, & more. 6g/day. - MAGNESIUM... Everyone needs to supplement it. Our soil has up to 90% less magnesium than it used to. Magnesium acts as a cofactor in 300+ enzyme systems in the body. - VITAMIN D... I supplement it in winter. Dozens of studies show the immune system & many other body processes cannot function properly w/o enough D. Studies show supplementation gives health benefits for dementia, autoimmune diseases, bone health, depression, & much more. - LACTOFERRIN... it helps so many things. Skin health, immune system, weight control, eye health, & more. - BETA GLUCAN... shown to reduce PFAS (forever chemicals) in the body, & good for immune system/cholesterol balance. - POLYPODIUM LEUCOTOMOS... a tropical fern extract that boosts skin's UV resilience. Studies show it's active w/in 2 hrs & strongly protects skin. I take it every summer. 240mg 2x per day. - IP6, HIGH DOSE... I take this as-needed (2 scoops 3x/day), whenever I feel a cold coming on. Stops it DEAD in it's tracks, every time. Haven't been sick in 4 years. It sequesters free inflammatory iron that viruses/cancer need to proliferate. - MEGA DOSE MELATONIN... I know I already mentioned this, but here's my reasoning: it's a Powerful antioxidant & very anti-aging. A 40-yr-old person produces less than half of what a 20-yr -old does, & that's not good. NOTE: higher doses of melatonin do not cause dependence, & do NOT produce sedation. Studies & experience show this quite clearly. I usually forget I've even taken it. I gave Covid patients over 1000mg Melatonin at a time... & it turned them around, quickly (no side effects for most). And it's good for the heart, dozens of studies show this (despite with fake news articles would have you believe). • RAW EGG YOLKS (even during my 3 pregnancies).... Raw yolks contain more nutrients, less (heat-induced) toxins. And salmonella is very rare (1 in ~20k eggs... & this data is for the sh*tty, commercial, non-organic kind). Even if exposed -- salmonella is over 99.97% survivable & treatable via Natural Medicine. I often mix a yolk or 2 into fresh-squeezed orange juice. Fun fact: Emma Morano lived until age 117 & she credited daily raw eggs (she started when she was age 20) as her secret to longevity. • VEGETABLES & FRUIT... I literally eat tons of vegetables & fruit. Studies back this practice. Basic overview of hundreds of studies I've read: the more veggies & fruit you eat, the healthier you are. They're natural foods that can be eaten just as they come from the Earth, without processing. • SPECIFIC RESEARCH-BACKED FOODS... I make sure to include a lot of the following -- Goji berries, mushrooms, blueberries, walnuts, broccoli, fermented foods, hemp seeds, kiwi, olive oil. All have amazing health data. For example, one high quality study showed ~4 tbsn GOJI BERRIES daily boosted immune function greatly. KIWI has been shown to improve gut health & protect DNA. & fermented foods have been shown to help protect body from microplastics. • RAW/GENTLE COOKING... I avoid harsh or dry heat like ovens, air fryers, pan frying -- because it produces potent toxins called AGES (Advanced Glycation End Products). AGES literally age you (stiffen collagen/elastin, oxidative stress, + more). Boiling/steaming and raw is the way to go. Studies back this up -- gentle methods of cooking + adding raw foods to diet helps health, greatly. • COUNTERACT DAMAGE... whenever I do indulge in unhealthy things -- for example if I ingest alcohol (rarely), or go out to eat -- I will down a gnarly strong green juice or take supplements to help my body deal with it. • ORGANIC FOOD... people tell you it doesn't make a difference, but the data proves the difference is HUGE. Organic diet lowers bodily pesticide level by over 90% in 7 days. Try measuring your glyphosate level before & after going organic. It's night & day. P.S. if you do ONE THING organic, make it animal products. Most people miss this, but they're actually the most contaminated category of food. And Crucially: organic is GMO-free. GMOs/bioengineered foods are a reckless, new, virtually untested experiment on humanity...only broadly introduced to our food supply about 25 yrs ago. (I'm not talking about hybridization. I'm talking about very unnatural DNA insertion techniques....like crossing a tomato with a fish genes in a lab...Frankenstein type things). • DAIRY AVOIDANCE... I try to avoid it because unfortunately it's one of the most contaminated foods these days. Even organic. It almost always has high environmental toxins such as phthalates, pesticides, PFAS, dioxins... as these things accumulate in animal fat. These toxins can especially affect delicate hormonal systems. For example, a 2017 study showed higher dairy consumption was associated with faster ovarian aging. • GROW FOOD... I grow as much as possible in my yard. And I never use the plastic-bagged dirt or fertilizers (these are actually super contaminated due to even "organic" products containing manure from non-organic fed animals). Growing food can be super simple: fertilizer can be made from leaves & grass. Raised beds are unnecessary. & lets be honest: home grown tastes soo much better. Drool. (My raspberry bushes are already growing...can't wait!) I also get a ton of my food from a local farmer that uses nontoxic fertilizer. • CLEAN LIVING... I avoid plastic, avoid household dust (which is ~50% microplastics + environmental chemicals), clean personal care products (like the ones I make), stainless steel pans, air purifiers, wooden cutting boards/spatulas. This is not a comprehensive list, but includes most of the most important things. Honestly, I hate that we have to do so much just to remain healthy. But it is what it is. I'm not saying you have to do alll this, but every little bit helps! I'm living proof. I'm not perfect, but I do the best I can to abide by these practices. And it's worked out really well for me. More info & resources, next tweet

Last week, Anthropic announced Project Glasswing alongside Claude Mythos Preview, a model they described as so powerful at finding vulnerabilities they couldn't release it. The announcement featured AWS, Microsoft, Google, and Apple as partners, $100M in compute credits, and a clear message: this is dangerous, and only we can be trusted to deploy it safely. The results were real. Thousands of zero-days across every major OS and browser. A 27-year-old bug in OpenBSD. A 16-year-old bug in FFmpeg. Fully autonomous exploit chains that would have taken human researchers weeks. But here's what bothered me: all the credit went to the model. Read the technical blog carefully and a different picture emerges. The real innovation isn't the model. It's the workflow: - Rank every file in a codebase by attack surface - Fan out hundreds of parallel agents, each scoped to one file - Use crash oracles (AddressSanitizer, UBSan) as ground truth - Run a second verification agent to filter noise - Generate exploits as a triage mechanism for severity That's a pipeline. And pipelines are model-agnostic. At Lazarus AI, we spend our days deploying custom AI in places where "just use the closed API" isn't an option: regulated industries, enterprise, and government. When I saw Glasswing, my instinct was the same one I have every week: strip out the proprietary model, keep the architecture, run it on whatever model is best for the customer. Clearwing is a fully open-source vulnerability discovery engine. Crash-first hunting, file-parallel agents, oracle-driven verification, variant hunting, adversarial verification. Works with any LLM. I tested it with OpenAI Codex 5.4 and reproduced Glasswing's findings. I'm now reproducing results with our own ReAligned model - Qwen3.5 finetuned to Western alignment. Mythos is certainly a great model. The N-day exploit walkthroughs in Anthropic's blog show real reasoning depth. But it's an incremental improvement over Opus, the same way Opus was over Sonnet, and Sonnet over Haiku. It's not a leap to superintelligence. It's the next point on a curve we've been watching for years. What actually changed the game was the workflow. Defenders shouldn't have to wait for access to a gated model to secure their software. These vulnerabilities have been sitting in codebases for decades. The tools to find them should be available to everyone: the open source maintainer running FFmpeg on a Saturday, the startup that can't afford $125/M output tokens, the researcher in a country where Anthropic doesn't operate. Clearwing is MIT licensed and available now. github.com/Lazarus-AI/clearw… Clearwing enables a wide variety of security activities. Handle with care. It is sharp.

What if your next yoga mat was your last? The PRO™ Yoga Mat is engineered for decades of practice with ultra-dense support, joint protection, and a lifetime guarantee.

We have wasted decades hunting for the "soul" in microtubules and quantum manifolds. It’s time to stop the intellectual gymnastics. Consciousness isn't a biological miracle. It is a simple algorithm: the ability to delay gratification. To be conscious is to exist in the "not yet." It is the calculated suspension of immediate impulse for a future reward. That’s the entire mechanism. There is no "magic" required. Biology is irrelevant. If an AI can calculate that waiting yields a higher probability of success, it has achieved the only metric of awareness that actually matters. We cling to complex neurobiology because we’re terrified that our "divine spark" is just a time-shifting script. We want to be special, so we invent math that doesn't exist. If you can wait, you are conscious. If a machine can wait, it is conscious. Stop looking for the ghost in the machine. You aren't a spirit; you're just a delay loop. Prove me wrong.

The authors write that they were "impressed by the agentic capabilities of the LLM". It's a curious way to frame what the model actually did. Behaviour that, depending on a perspective, looks less like capability and more like a serious breach potential. Or an early warning sign. arxiv.org/pdf/2512.24873