
"Humans [are going to] become more and more valuable because they will be the long pole in the tent to progress - and that progress is going to accelerate and get faster and more robust." - Travis Kalanick
T Ay.
3.5K posts

@ayedtay
ml x data @runwayml | prev. photoroom - - - opinions and rants 100% my own

"Humans [are going to] become more and more valuable because they will be the long pole in the tent to progress - and that progress is going to accelerate and get faster and more robust." - Travis Kalanick


The paper I’ve been most obsessed with lately is finally out: nbcnews.com/tech/tech-news…! Check out this beautiful plot: it shows how much LLMs distort human writing when making edits, compared to how humans would revise the same content. We take a dataset of human-written essays from 2021, before the release of ChatGPT. We compare how people revise draft v1 -> v2 given expert feedback, with how an LLM revises the same v1 given the same feedback. This enables a counterfactual comparison: how much does the LLM alter the essay compared to what the human was originally intending to write? We find LLMs consistently induce massive distortions, even changing the actual meaning and conclusions argued for.


The paper I’ve been most obsessed with lately is finally out: nbcnews.com/tech/tech-news…! Check out this beautiful plot: it shows how much LLMs distort human writing when making edits, compared to how humans would revise the same content. We take a dataset of human-written essays from 2021, before the release of ChatGPT. We compare how people revise draft v1 -> v2 given expert feedback, with how an LLM revises the same v1 given the same feedback. This enables a counterfactual comparison: how much does the LLM alter the essay compared to what the human was originally intending to write? We find LLMs consistently induce massive distortions, even changing the actual meaning and conclusions argued for.


please don’t speak on behalf of the industry - this is beyond 🤦


Let me explain exactly why every new subdivision in America looks like the top photo, because the math is wild. A mature tree increases a home's value by 7 to 19 percent. On a $400,000 house, that's $28,000 to $76,000. A single shade tree produces the cooling equivalent of ten room-size air conditioners running 20 hours a day. One tree on the west side of a house cuts energy bills by 12 percent within 15 years. The bottom photo is worth more, costs less to live in, and sells faster. This has been documented by the University of Washington, Clemson, Michigan State, and the USDA. The data is not in dispute. Removing those trees saves the builder roughly $5,000 per lot. Concrete trucks need twice the dripline radius of every standing tree. Utility trenches need flat ground. A bulldozer flattens 200 lots in an afternoon. Preserving trees adds weeks and thousands per home. So the developer pockets $5,000 in savings and the buyer eats $50,000 in lost value for the next two decades. The person making the decision and the person paying for it have never been in the same room. The Woodlands, Texas is the proof of what happens when they are. George Mitchell bought 28,000 acres of Houston timberland in 1974 and preserved 28% as permanent green space. He forced McDonald's to build behind the tree canopy. That McDonald's became one of the highest-volume locations in Texas. The first office building, designed to reflect the surrounding forest so you couldn't see it from the street, leased completely. The Woodlands median home price today: $615,000. Katy, a comparable Houston suburb that clear-cut: $375,000. Named #1 community to live in America two years running. Fifty years of data. The trees are worth more than removing them saves. Developers clear-cut anyway because they sell the house once and leave. You live in it for 30 years.




A breakthrough in real-time video generation. As a research preview developed with @NVIDIA and shared at @NVIDIAGTC this week, we trained a new real-time video model running on Vera Rubin. HD videos generate instantly, with time-to-first-frame under 100ms. Unlocking an entirely new creative paradigm and bolstering the foundations of our General World Model, GWM-1. Real-time generation opens a fundamentally different design space for video models and world simulation. We're investing in co-designing our models alongside advances in hardware to keep pushing this frontier.

During last year's Paris AI Action Summit, Fluidstack announced plans for a €10 billion AI data center championed by president @EmmanuelMacron. A year on, and Fluidstack has pulled out of the project. Scoop from @BenoitBerthelot bloomberg.com/news/articles/…


BREAKING: Kim Jong-un has won North Korea's parliamentary elections with 99.93% of the vote.



Sorry to be the downer because this is an impressive story in some senses. But it is ~trivially easy to make a single mRNA vaccine. It's not hard. I cure mice of various cancers with various therapeutics all the time. I've made mice lose more weight in a month than tirzepatide does in a year. What is hard and expensive is proving its BOTH safe AND effective **in a randomized and controlled study in humans** while ALSO manufacturing it at clinical scale and grade. I am happy for this man and his dog. It is impressive. But y'all are overhyping it.


This is wild. theaustralian.com.au/business/techn…

"Humans [are going to] become more and more valuable because they will be the long pole in the tent to progress - and that progress is going to accelerate and get faster and more robust." - Travis Kalanick