Alistair

@godofprompt @grok how does this compare to using Grok's web search tools? How does Grok have an edge over Claude in this area?

🚨 BREAKING: Every prompt you're using right now is built on what worked months ago. Someone just open-sourced the solution. A Claude Code skill that scans Reddit and X from the last 30 days on any topic you give it, then writes you a fully structured, copy-paste-ready prompt based on what the community has actually figured out. Not last quarter. Not last year. This month. You type "/last30days prompting techniques for ChatGPT for legal questions" and it returns the exact patterns real lawyers and power users are running right now. Complete with a deployment-ready prompt you drop in and use immediately. It works for anything: → Midjourney techniques the Discord discovered this week → Cursor rules that actually work in 2026 → Claude prompting patterns from builders shipping real products → Suno, Runway, code generation, whatever is moving right now This is what prompt engineering looks like when it stops being a static document and starts being live intelligence. The entire AI prompting landscape updates faster than any guide, course, or thread can keep up with. This skill treats that reality as a feature, not a problem. Most people are still Googling for prompts that got patched out three updates ago. This pulls from the source: real users solving real problems in real time. 100% open source. MIT License. The gap between people using fresh prompts and people recycling old ones is about to get very visible.

@SamaHoole @evolvee33 Just measure your retinol to get a baseline and continue to monitor...

@evolvee33 Be careful consuming liver for too long. You can definitely overload on Vitamin A.

Raw kidney beans: five of them will hospitalise you. Raw cassava: kills people regularly in food-insecure regions if not processed correctly. Raw elderberries: cyanide compounds. Toxic without cooking. Raw potatoes: solanine. Neurological symptoms in quantity. Raw grain: mycotoxins, lectins, phytates. Requires processing to be safely edible. Raw spinach: oxalates that accumulate in soft tissue with regular use. Raw ribeye: safe. Eaten raw across cultures throughout history. No processing required. The plant kingdom is defending itself. This is not metaphor. The chemistry is real and documented. You don't have to cook a steak to avoid hospital. You absolutely have to cook a red kidney bean.

@LeCodeBusiness @HansAmato Agreed, can't really call it an experiment if you didn't systematically modify once behaviour at a time and retest T at each stage. To strengthen the anecdote (that's all it is), you'd need the client to return back to his initial diet and lifestyle and retest to see if T drops.

@HansAmato Activity, diet, and stress load all changed simultaneously. You can't attribute a 300-to-1000 jump to any single intervention. This is a compelling story, not a replicable protocol.

Had a client doing 30,000 steps a day. MMA. Lifting. Strict carnivore. Zero rest days. Looked like the healthiest person in any room His testosterone was 300 Every doctor would have looked at his lifestyle and said "this makes no sense." Because the assumption is that healthy habits produce healthy hormones They don't. Not when output exceeds input by that much His body was hemorrhaging energy faster than he could replace it. Carnivore with no carbs meant his thyroid had downregulated. The insane training volume meant cortisol was permanently elevated. His body was cannibalizing itself to keep the engine running We cut his activity. Added carbs back. Increased nutrient density with liver and organ meats. Lowered the total stress load on his system His testosterone went from 300 to 1000 Not 500. Not 700. One thousand The "hardest worker in the room" mentality is destroying men's hormones. More is not better when your body doesn't have the fuel to recover from it. Every rep without adequate nutrition is a withdrawal from a bank account you're not depositing into If you're training hard, eating "clean," doing everything right, and still feel like garbage with no sex drive and no energy. You might be the fittest man in the room running on the hormones of a 75 year old Eat more. Do less. Watch the number climb I go deep on all of this on Substack. Gut protocols, hormones, what most "clean eaters" are getting wrong. Link in bio

@drjasonfung Whey is a complete protein that scores more highly on DIAAS than beef. Don't see what the issue is. It's convenient to consume.

Whey was primarily fed to pigs until recently, when people (marketing geniuses) figured out you could instead feed it to humans as protein powders, make health claims and 100x your money. (eg. Starbucks protein coffee). If you want to eat whey, go get some Norwegian brown cheese.

@SamaHoole How would you recommend experimenting, remove it completely? Gradually reduce?

I haven't used salt in two years. I want to be careful here, because salt is one of those things where the carnivore community has quietly developed a strong collective position that I think deserves examination. Salt is not an animal product. It does not appear in the archaeological record of Palaeolithic diet in any meaningful quantity. It is a mineral additive that humans adopted as a preservation technology, to stop meat rotting, not because the physiology required it. I cut it out two years ago, and the following resolved within weeks: post-workout cramps, fast heartbeat in the evening, low-level anxiety that I'd attributed to other causes, and waking at 2am to use the bathroom. All gone. The mechanism, as I understand it: sodium and water maintain a feedback loop. More sodium requires more water to maintain osmotic balance. More water increases blood volume. Increased blood volume at rest, when you're not actively losing electrolytes through exercise, puts unnecessary load on the cardiovascular system and disrupts the signalling that regulates sleep and urine production. Carnivore resolves so many things. Salt quietly undoes some of them. Worth experimenting with.

@tetsuoai @SecWar Who tells the AI what the truth is?

@SecWar Truth seeking AI is the safest path towards AGI

This week, Anthropic delivered a master class in arrogance and betrayal as well as a textbook case of how not to do business with the United States Government or the Pentagon. Our position has never wavered and will never waver: the Department of War must have full, unrestricted access to Anthropic’s models for every LAWFUL purpose in defense of the Republic. Instead, @AnthropicAI and its CEO @DarioAmodei, have chosen duplicity. Cloaked in the sanctimonious rhetoric of “effective altruism,” they have attempted to strong-arm the United States military into submission - a cowardly act of corporate virtue-signaling that places Silicon Valley ideology above American lives. The Terms of Service of Anthropic’s defective altruism will never outweigh the safety, the readiness, or the lives of American troops on the battlefield. Their true objective is unmistakable: to seize veto power over the operational decisions of the United States military. That is unacceptable. As President Trump stated on Truth Social, the Commander-in-Chief and the American people alone will determine the destiny of our armed forces, not unelected tech executives. Anthropic’s stance is fundamentally incompatible with American principles. Their relationship with the United States Armed Forces and the Federal Government has therefore been permanently altered. In conjunction with the President's directive for the Federal Government to cease all use of Anthropic's technology, I am directing the Department of War to designate Anthropic a Supply-Chain Risk to National Security. Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic. Anthropic will continue to provide the Department of War its services for a period of no more than six months to allow for a seamless transition to a better and more patriotic service. America’s warfighters will never be held hostage by the ideological whims of Big Tech. This decision is final.

@DaveShapi What Claude model are you using? Sonnet/Opus?

Grok has become my daily driver but it still glazes too much, especially on longer conversations. It feels VERY high effort, much higher effort than Claude, Gemini, and ChatGPT. I just asked ChatGPT PRO to help design a hypothesis for an experiment and it completely ignored the instructions. It spent 25 minutes "thinking" and rendered a "here's how you might consider going about designing such an experiment." No testable hypothesis. When I pushed, it made up a bunch of shit about being cautious and avoiding over-confidence (my dude, that's the point of a hypothesis). With literally the same exact prompt, Grok did the literature review and rendered some hypothetical gradients (with specifics) in a few seconds. Claude... well it tried. I let it think for more than an hour but I guess it froze or something. After about 600 sources, it just froze. And anyways, even when it doesn't freeze, Claude tends to sandbag, hedge, and move the goalposts just as bad as ChatGPT. Gemini is the second highest effort AI, in that it pretty much never feigns ignorance, disinterest, or misunderstanding (seriously, we've arrived at a world where Claude and ChatGPT literally feign disinterest or misunderstanding to avoid doing work!) With that being said, Gemini still hallucinates WAY TOO MUCH. Even 3.1 pro. I frequently have to catch it and admonish and say "yeah, that thing you said, which is actually really important, is not actually true." Fortunately, Gemini has zero ego and pushback. It still has that "Wow! You're exactly right! Good catch!" energy (it feels like a Golden Retriever, honestly). But Gemini's hallucination rate makes it unusable for serious work. Grok, so far, does not seem to hallucinate nearly as much, and generally not on big things. I will say, however, that Grok tends to get itself into search bubbles i.e. it doesn't stop and think "hey maybe I need to broaden my search" without being prompted. With that being said, it does do a good job if you explicitly say "let's steelman this" - it is very happy to look for countervailing evidence. But I shouldn't have to prompt it. But still, when you pay $100 to $200 per month for an AI, you should NEVER have to beg it to be useful. And when you chastise the model for being useful it should NEVER be defensive. ChatGPT and Claude have become utterly useless for serious work due to this. Gemini and Grok are the kings of frontier research as far as I'm concerned. Anything that requires speculation, forecasting, modeling, or predicting, do NOT use Claude or ChatGPT. They are fully and unforgivably lobotomized on those kinds of tasks. In some cases, they are hyper sensitive to your opening framing. But even when I try and think "what's the best way to frame this query without upsetting the AI?" it rarely works today. As many of you in the comments have pointed out, using some of these chatbots is like babysitting the AI. This is why Grok and Gemini have far superior UX - if they get it wrong, they don't get hung up on pushback or criticism. They do an about-face and will completely change their approach. ChatGPT and Claude get irrevocably stuck in a rut, so you have to abandon the conversation and start from scratch.

Grok 4.20 passes the Caitlyn Jenner AI Test. ChatGPT fails. Gemini fails.

@cvByJd @EightBitElon @claudeai @elonmusk @xai Grok for heavy lifting --> Claude for final polish/validation. Seems to be a good sweet spot. Never rely on one LLM for an entire workflow.

@EightBitElon @claudeai @elonmusk and @xai is better in my opinion for a lot of real life use cases

This is Claude Sonnet 4.6: our most capable Sonnet model yet. It’s a full upgrade across coding, computer use, long-context reasoning, agent planning, knowledge work, and design. It also features a 1M token context window in beta.

@Official1Rsl @claudeai The "cool down" periods on Claude are outrageous. I've had to wait multiple days to reset my usage limits on a pro subscription.

@claudeai It is outrageous and downright humiliating that the conversation gets shut down before I even finish speaking, and that I’m not given a clear explanation for such an absurdly long limit period.”

We're bringing some of Claude’s most-used features to the free plan. File creation, connectors, and skills are all now available without a subscription.

@ellymelly Eat whole, real foods and avoid anything that didn't exist 200 years ago.

Help me settle a completely non-political debate. I have this friend. He takes health and fitness super seriously. He loves all those protein powders, energy bars, any snack that has the word 'protein' on it basically. You know what I'm talking about. I - won't touch any of it. Ever. To me, it looks like it was cooked up in a pharmacy aka NOT FOOD. I'd rather have the odd cupcake. Am I being too harsh? Where are my fitness bros at.

@helios_brah @grok Is there scientific evidence that the evolutionary reduction in human brain size (vs. Neanderthals or early Homo sapiens) correlates with lower intelligence? Curious what the data shows.

So basically we either lost 150 grams of brain tissue (from when we were Neanderthals) by moving inland, away from DHA rich seafood (our brain is 60% DHA) Or we lost it because polarised artificial light (yes, fire) effectively gave us constant Traumatic Brain Injury (TBI) so that we shrunk brain and eye size to cope Or both Either way I will be eating my oysters and avoiding light at night mfers ✌️

@Cobratate Yes, but he still died... Like everyone else. The great equalizer.

@NextGenVisionar @MarioNawfal Thanks ChatGPT.

@MarioNawfal Trump eyeing Greenland after Venezuela win? Generals shut it down hard—illegal without Congress. But securing Arctic resources & strategic edge… bold vision or wild idea? Either way, he’s thinking big.

🚨🇺🇸🇬🇱 TRUMP: "MAYBE WE'LL INVADE GREENLAND," U.S. TOP COMMANDERS: "YOU CAN'T BE SERIOUS" Apparently, Trump asked special forces to draw up an invasion plan for Greenland after getting pumped from the Venezuela operation. But top generals pushed back hard, saying it’s illegal without Congress and "not happening." Source: Daily Mail

@hubermanlab @HHSGov @hubermanlab when are you going to sit down again with @DrJackKruse and hash it out?

The new @HHSGov food pyramid’s take on veganism and grains.

@OffgridIreland @TasteAtlas @grok @grok under new testament scripture, does the Bible restrict any foods?

@TasteAtlas @Grok What's the bibles take on these

Blood sausages are among the oldest known sausages in the world. They are made by mixing animal blood with fat, grains or bread, and mild spices, then gently cooked until set. Found across Europe and beyond, they developed as a way to preserve nutrients after slaughter and to avoid waste. Rich in iron and calories, they were especially valued in winter, when fresh food was scarce and energy-dense meals were needed.

@9mmsmg Firefox + ublock + video background play (Android). Not sure if this works with iOS.

I can't handle ads anymore, they drive me insane. I don't use any of the premium + features on X and last year I downgraded to regular premium. It took me one week to upgrade again. The experience was awful and it made the app miserable. I abhor YouTube/Google but I still begrudgingly subscribe to premium because it's totally unusable with ads. I exclusively use both X and YT on the app on phones and tablets and I hate using the websites. I don't know how some people do it.

@bitcoinand_beef It would be good to see some more spectrographs of household halogen/incandescent lighting. In Australia we have bathroom heat lamps, but I'm unsure if they would contain any UVA.

High wattage halogen lighting is likely the best daytime indoor light source. 3000K, extremely bright, tons of infrared. This 1000 watt lamp costs $60

@AshPolitik And Donald Trump will never be POTUS...

Albo won't resign Albo will win the next election, 100 seats Sussan Ley will lose the leadership Pauline, Beetroot and One Nation will never be in government Murdoch press will continue to bleed readers and revenue. Happy New Year!

@louisanicola_ Tldr; your training regiment should be tailored to your specific fitness/health goals.

Busting Muscle Myths: What Science REALLY Says Think muscle growth is all about heavy weights and genetics? The science says otherwise Here’s what research-backed evidence reveals: 1️. Heavy weights are NOT required to build muscle • Light weights can grow muscle just as well • Key factor: training with high effort, close to fatigue • Great news for beginners, rehab, and injury recovery 2. Time off doesn’t erase your progress • Muscle shrinks, but your body remembers • Past training leaves epigenetic “muscle memory” • Results: faster muscle regain when you return 3. Muscle fiber type is NOT fixed at birth • Endurance training shifts fibers toward slow, efficient types • Power training shifts fibers toward fast, explosive types • Lifestyle can dramatically reshape muscle makeup Your muscles are adaptable, intelligent, and trainable at ANY stage of life. Train smarter, not heavier. Your muscles are listening