Jason Hickey

Interested in AI, weather, climate, math, sculpture, art, glass, meaning. Views here are personal and not necessarily those of my employer.

[Jason Hickey @bapadadada]

Beautiful find — congratulations! It's now machine-verified: I formalized the conjecture as stated in the literature (arXiv:2308.02651, Conjecture 1.3) and checked the counterexample in Lean 4 — every routing, path-completeness included, down to three standard axioms.

Discovered by GPT-5.6, verified with Claude. The kernel is the referee.
https://t.co/Ta8DZtgUzg

[Jason Hickey @bapadadada]

An 87-year-old conjecture fell this weekend — and the counterexample is now machine-verified. det Jac ≡ −2 and the three-point fiber, formally checked in Lean 4 + mathlib, sorry-free, axioms = [propext, Classical.choice, Quot.sound]. One self-contained file: https://t.co/6el2GAtQcf… #Lean4 #math

[levent @__alpoge__]

hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final

((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)

[Jason Hickey @bapadadada]

Independent formal verification of the counterexample as stated in the preprint: det Jac F ≡ −2 (formal partial derivatives) + the full three-point fiber, over ℚ and ℂ. Lean 4 + mathlib, one self-contained file, sorry-free, axioms = [propext, Classical.choice, Quot.sound]. https://t.co/6el2GAtQcf

[Jimmy Ba @jimmybajimmyba]

Last day at xAI.

xAI's mission is push humanity up the Kardashev tech tree. Grateful to have helped cofound at the start. And enormous thanks to @elonmusk for bringing us together on this incredible journey. So proud of what the xAI team has done and will continue to stay close as a friend of the team. Thank you all for the grind together. The people and camaraderie are the real treasures at this place.

We are heading to an age of 100x productivity with the right tools. Recursive self improvement loops likely go live in the next 12mo. It’s time to recalibrate my gradient on the big picture. 2026 is gonna be insane and likely the busiest (and most consequential) year for the future of our species.

[Jeff Dean @JeffDean]

Nice example of the multimodal OCR capabilities of Gemini 2.5 Pro. Just give it dusty old records and ask for JSON!

https://t.co/gTBEXkGLYB

"Which is almost perfect. It has misread a small number of the hard-to-read 8s as 5s. But it is spot-on apart from that. No hallucination.

A revolution is here. I think we will soon be able to rescue historical weather observations data much more efficiently and quickly."

[Jeff Dean @JeffDean]

Great to see AI-based weather prediction being rolled out across Africa. This is important because traditional weather forecasts rely much more on ground-based radar, and in Africa, there simply aren't enough to give high quality traditional forecasts.

https://t.co/yonBR65ESU

[Peter Battaglia @PeterWBattaglia]

Today we’re thrilled to announce that real-time and historical AI-based weather forecasts from @Google’s WeatherNext suite of models are now available on Earth Engine and BigQuery. Anyone can access and use these data for research, analysis and operational decision making, which we hope will help accelerate research and development in the weather and climate community.

You can check out the AI-powered forecasts here: https://t.co/Dj1lq1VyZ4

[Google AI @GoogleAI]

We’re putting our most advanced AI for weather forecasting into more people’s hands. ☁️🌐

Scientists can now access @Google's WeatherNext models as well as real-time & historical data to accelerate research, improve how we respond to disasters, & more →https://t.co/M3TvXO5Q1v

[Jason Hickey @bapadadada]

We are incredibly excited to launch the Open Buildings 2.5D Temporal Dataset, open-sourcing annual data about building presence and heights across the Global South from 2016-2023.

[Jeff Dean @JeffDean]

An exciting update to our Open Buildings dataset. It now features building presence over time (as well as building height information). It does this by combining high resolution (50cm) but infrequent satellite images and low resolution (1000cm) but frequent (every 5 days). It uses the high resolution imagery as a teacher to predict the higher resolution version from the more frequent lower resolution imagery.

This enables smooth analysis of buildings over time, as you can see in time lapse below of buildings and roads being constructed on the outskirts of Kumasi, Ghana.

Potential uses include:

Government agencies: Gain insights into urban growth patterns to inform planning decisions

Humanitarian organizations: Quickly assess the extent of built-up areas in disaster-stricken regions

Researchers: Track development trends, and study the impact of urbanization on the environment.

Read more on the blog post:
https://t.co/XYaMCOvUdR

Visualize different spots on Earth or count buildings in Earth Engine:
https://t.co/Ccb8CV5h0N

Dataset:
https://t.co/RZogn15KfJ

Kudos to the whole @GoogleResearch team that worked on this! 🎊
Wojciech Sirko, Krishna Sapkota, Juliana T. C. Marcos, Abigail Annkah, Olivia Graham, Emmanuel Asiedu Brempong, Abel Korme, Mohammed Alewi Hassen, Sella Nevo, Tomer Shekel, Abdoulaye Diack, Jason Hickey, and John Quinn.

[Stephan Hoyer @shoyer]

I'm incredibly proud to share NeuralGCM, our new AI and physics based approach to weather and climate modeling with state-of-the-art accuracy, published today in @Nature:
https://t.co/rtQa0Lq0fT

[Jason Hickey @bapadadada]

We just had an amazing visit at the World Food Program (WFP) on the theme of how AI can be used for humanitarian purposes. I was impressed by the scope of work at WFP, and by their drive to have real and immediate impact. Our work at Google Research Africa was well received.

[ECMWF @ECMWF]

#MachineLearning is playing an increasing role in #WeatherForecasting.

Find out about machine-learning-based progress in improving the initial conditions and the trajectory of physics-based forecasts
➡️https://t.co/6d1b4iiC25

[Sean O' Riordan @seanorphoto]

Absolutely biblical skies in Tasmania at 4am this morning. I’m leaving today and knew I could not pass up this opportunity for such a large solar storm. Here’s the image. I actually had to de-saturate the colours. Clouds glowing red. Insane. Shot on Nikon. Rt appreciated

[Google AI @GoogleAI]

Meet @GreyNearing, a Research Scientist working on flood forecasting at Google Research! Grey & his team produce real-time flood forecasts & alerts in 80+ countries through the Flood Hub to help keep people safe → https://t.co/6GrBUBKWjQ
Learn more → https://t.co/jFXNLM5rib

[Daniel Rothenberg @danrothenberg]

Their marquee result is Figure 1, which has a shocking punchline: in poorer countries, there's a huge gap in forecast quality, so much so that the median 1-day forecast is only as accurate as a 7-day forecast in richer countries! 3/n

[NWS Bay Area 🌉 @NWSBayArea]

Extremely Dangerous Winds.

The waters from the Big Sur Coast to 60 nautical miles out have been upgraded to Hurricane Force Wind Warnings for winds 35 to 55 kt with gusts up to 80 knots (roughly 92 mph).

#CAwx

[Jason Hickey @bapadadada]

We're visiting the Zambia Meteorological Department for the WISER Early Warnings for Southern Africa live testbed, really exciting https://t.co/8OUd5KPUTN

[African Institute for Mathematical Sciences (AIMS) @AIMS__Network]

Our call for #applications for the MSc in Mathematical Sciences is open!

Are you ready to embark on a #Transformative academic journey that will elevate your expertise in #MathematicalSciences? Apply via https://t.co/QACa43C1jo by 29/03/2024

#AIMS2024 #Opportunity #FullyFunded

[Jason Hickey @bapadadada]

Congratulations to the GraphCast team for spectacular results using AI for medium-range weather forecasting! The wind is changing!

[Google DeepMind @GoogleDeepMind]

Presenting GraphCast: our state-of-the-art AI model delivering 10-day weather forecasts with unprecedented accuracy in under one minute. 🌦️

It can even help predict the potential paths of cyclones further into the future.

Here's how it works. 🧵 https://t.co/ygughpkdeP