無敵 光線 尊厳

When did fetal alcohol syndrome become a fashion statement?

Is this the best use of AlphaFold to date? ;)

Narrative scrambler: Labour share of income has fallen in many countries, but not the United Kingdom.

AI deradicalises people...More AI less Social Mediea WTF!

@travelingflying *larger homes in outer boroughs

@travelingflying Housing and Economic Aspiration: Many white working-class families left the inner city during the 20th-century slum clearance programs, moving to new, larger boroughs of London and then selling right to buy and moving out of the city.

Bill Maher: ”In fifty years, London went from 86% White to 36% White. This is great, it’s a happy fact. London is now mostly people of color, and I’m applauding it.’” Meanwhile, London: 1960 98% White 0.01 Knife crime per 1,000 2026 36% White 1.82 Knife crime per 1,000

@travelingflying 53.8% White (36.8% White British rest White Other), 20.7% Asian, 13.5% Black, 5.7% Mixed, and 6.3% Other.

@mike_lustgarten 3/System Guardrails & Agents: Implement automated checks that compare the output against the source material and use "critics" or multiple models to cross-verify answers.

Working with LLMs to figure out solutions to hard problems is an upgrade over traditional search engines, but man, (currently) time consuming, as they need to be continuously challenged to make sure it's not a hallucination I worry about people who get the answer immediately, and then go away without any real understanding...

@mike_lustgarten 2.2/Contextual Constraints: Instruct the model: "Answer only using the provided context" or "If you do not know, say you do not know".

@mike_lustgarten 2.1/ Advanced Prompting Techniques:Chain-of-Thought (CoT): Ask the model to break down reasoning step-by-step, improving accuracy for complex tasks. Few-Shot Prompting: Provide examples within the prompt to guide the model on the expected format and level of accuracy.

@mike_lustgarten 1/ Retrieval-Augmented Generation (RAG): Connect the LLM to trusted, external data sources (databases, documents, search engines) to provide context before it answers, forcing it to rely on facts rather than internal training data.

Nemoclaw @Plinz

@mike_lustgarten Nemoclaw

I'm not yet using Agentic AIs to assist with biomarker data interpretation, but juggling info from standard LLMs (Chat, Grok, Gemini, Perplexity, Claude), I become the Agentic AI Anyone else using a similar approach, and is there a better way? I'm open to using a better approach...

@Msamalam The wider metropolitan area of Paris (aire d'attraction) is significantly larger, with a population surpassing 13 million people

I actually don’t understand how France has 69 million people. Where are they? 5 largest cities: Paris 2 million Marseille 0.8MM Lyon 0.5MM Toulouse 0.5Mm Nice 0.3MM Where are they all hiding? In the countryside? The “big” cities are so tiny.

@micahsb @ramez @stevehou @tszzl Not Luxembourg : "Public Attitude: Over 70% of citizens have used AI tools, with 77% taking a pragmatic view that recognizes both the benefits and risks"

Once we removed the two outliers the trend is much more striking. Lower income countries understand the opportunity AI present: an opportunity to siphon knowledge from the advanced economies and catch up! @tszzl had made this point persuasively. To the extent that AI is a shock to the high margin services sector, less advanced economies have smaller or little sophisticated services economies to begin with. They have much less to lose and plenty to gain by building out cheaper alternatives to chip away at expensive services solutions in the advanced economies. AI could be the next “China shock” but for services this time.

@uncle_deluge The hole in Ozone layer (which is healing) exposes fair skinned people to Extreme UV levels.

Then I wonder why the people in these two regions with the same ethnicity and basically the same diet have wildly different skin cancer rates

@tszzl The rest of the galaxy is left with 10% dimming due to the astrophages. Less luminosity would mean higher chance for more life in the galaxy (dimmer stars are longer lived). If 10% less is + or - depends on each solar system state at the time.

(hail mary spoilers so please skip if you care) isn’t the situation by the end that every single star in galaxy dies (or dims enough to kill all planetary life) except the two that are saved and the third where the taomeba originate

A periodic reminder to leave behind simplistic & naive notions.