What is it worth to know the weather forecast 1 day earlier? That’s not a hypothetical question, traditional forecasting systems have been improving their skill at a rate of 1 day per decade. In other words, today’s 6-day forecast is as accurate as the 5-day forecast ten years ago. No one expects this rate of improvement to hold steady, it has to slow down eventually, right? Well in the last couple years GPUs and modern deep learning have actually sped it up.
Since 2022 there has been a flurry of weather deep learning systems research at companies like NVIDIA, Google DeepMind, Huawei and Microsoft (some of them built by yours truly). These models have little to no built-in physics and learn to forecast purely from data. Astonishingly, this approach, done correctly, produces better forecasts than traditional simulations of the physics of our atmosphere.
Jayesh and Cris came face-to-face with this technology’s potential while they were respectively leading the [ClimaX](https://arxiv.org/abs/2301.10343) and [Aurora](https://arxiv.org/abs/2405.13063) projects at Microsoft. The foundation models they built improved on the ECMWF’s forecasts, considered the gold standard in weather prediction, while only using a fraction of the available training data. Our mission at Silurian is to scale these models to their full potential and push them to the limits of physical predictability. Ultimately, we aim to model all infrastructure that is impacted by weather including the energy grid, agriculture, logistics, and defense. Hence: simulate the Earth.
Before we do all that, this summer we’ve built our own foundation model, GFT (Generative Forecasting Transformer), a 1.5B parameter frontier model that simulates global weather up to 14 days ahead at approximately 11km resolution (https://www.ycombinator.com/launches/Lcz-silurian-simulate-t...). Despite the scarce amount of extreme weather data in historical records, we have seen that GFT is performing extremely well on predicting 2024 hurricane tracks (https://silurian.ai/posts/001/hurricane_tracks). You can play around with our hurricane forecasts at https://hurricanes2024.silurian.ai. We visualize these using [cambecc/earth] (https://github.com/cambecc/earth), one of our favorite open source weather visualization tools.
We’re excited to be launching here on HN and would love to hear what you think!
What else do you hope to simulate, if this becomes successful?
But it's non-trivial to scale these new techniques into the field. A major factor is the scale of interest. FEMA's FIRMaps are typically at a 10m resolution not 11km.
They’re selling height maps of South-Africa, primary for flooding prediction for insurance companies.
Smart & friendly bunch.
They're a cool little team based in Copenhagen. Would be useful, for example, to look at the correlation between your weather data and regional energy production (solar and wind). Next level would be models to predict national hydro storage, but that is a lot more complex.
My advice is to drop the grid itself to the bottom of the list, and I say this as someone who worked at a national grid operator as the primary grid analyst. You'll never get access to sufficient data, and your model will never be correct. You're better off starting from a national 'adequacy' level and working your way down based on information made available via market operators.
Signed,
A California Resident
It seems like this is another instance of The Bitter Lesson, no?
I thought this was a good quote:
> We want AI agents that can discover like we can, not which contain what we have discovered.
Deep Blue wasn't a brute-force search. It did rely on heuristics and human knowledge of the domain to prune search paths. We've always known we could brute-force search the entire space but weren't satisfied with waiting until the heat death of the universe for the chance at an answer.
The advances in machine learning do use various heuristics and techniques to solve particular engineering challenges in order to solve more general problems. It hasn't all come down to Moore's Law.. which stopped bearing large fruit some time ago.
However that still comes at a cost. It requires a lot of GPUs, land, energy, and fresh water, and Freon for cooling. We'd prefer to use less of these resources if possible while still getting answers in a reasonable amount of time.
It's certainly true that "just throw a bunch of GPUs at it" is wasteful, but it does achieve results.
Notably forecast skill is quantifiable, so we'd need to see a whole lot of forecast predictions using what is essentially the stochastic modelling (historical data) approach. Given the climate is steadily warming with all that implies in terms of water vapor feedback etc., it's reasonable to assume that historical data isn't that great a guide to future behavior, e.g. when you start having 'once every 500 year' floods every decade, that means the past is not a good guide to the future.
From the post.
What more did you want from them? (Genuine question.)
nullschool is obscure enough to the general audience that when I saw it there was an immediate red flag.
If only specialized scientists can see the difference between the sites, it's a presentation problem.
The biggest issue is that the basic data model for population behavior is a sparse metastable graph with many non-linearities. How to even represent these types of data models at scale is a set of open problem in computer science. Using existing "big data" platforms is completely intractable, they are incapable of expressing what is needed. These data models also tend to be quite large, 10s of PB at a bare minimum.
You cannot use population aggregates like census data. Doing so produces poor models that don't ground truth in practice for reasons that are generally understood. It requires having distinct behavioral models of every entity in the simulation i.e. a basic behavioral profile of every person. It is very difficult to get entity data sufficient to produce a usable model. Think privileged telemetry from mobile carrier backbones at country scales (which is a lot of data -- this can get into petabytes per day for large countries).
Current AI tech is famously bad at these types of problems. There is an entire set of open problems here around machine learning and analytic algorithms that you would need to research and develop. There is negligible literature around it. You can't just throw tensorflow or LLMs at the problem.
This is all doable in principle, it is just extremely difficult technically. I will say that if you can demonstrably address all of the practical and theoretical computer science problems at scale, gaining access to the required data becomes much less of a problem.
IMO the short answer is that such models can be made to generate realistic trajectories, but calibrating the model the specific trajectory of reality we inhabit requires knowledge of the current state of the world bordering on omniscience.
[0]: https://www.santafe.edu/research/results/working-papers/asse...
Specifically, I could imagine throwing current weather data at the model and asking it what it thinks the next most likely weather change is going to be. If it's accurate at all, then that could be done on any given day without further training.
The problems happen when you start throwing data at it that it wasn't trained on, so it'll be a cat and mouse game. But it's one I think the cat can win, if it's persistent enough.
One nit on your framing: NeuralGCM (https://www.nature.com/articles/s41586-024-07744-y), built by my team at Google, is currently at the top of the WeatherBench leaderboard and actually builds in lots of physics :).
We would love to metrics from your model in WeatherBench for comparison. When/if you have that, please do reach out.
Re NeuralGCM, indeed, our post should have said "*most* of these models". Definitely proves that combining ML and physics models can work really well. Thanks for your comments!
Main takeaway, gives me some hope:
Our results provide strong evidence for the disputed hypothesis that learning to predict short-term weather is an effective way to tune parameterizations for climate. NeuralGCM models trained on 72-hour forecasts are capable of realistic multi-year simulation. When provided with historical SSTs, they capture essential atmospheric dynamics such as seasonal circulation, monsoons and tropical cyclones.
In recent years, computing has both expanded as a field and grown in its importance to society. Similarly, the research conducted at Google has broadened dramatically, becoming more important than ever to our mission. As such, our research philosophy has become more expansive than the hybrid approach to research we described in our CACM article six years ago and now incorporates a substantial amount of open-ended, long-term research driven more by scientific curiosity than current product needs.
(Former Google employee, but I have no inside knowledge; this is just my speculation from public data.)
Owning your own data and serving systems can also make previously impossible features possible. When I was a Google intern in 2007 I attended a presentation by someone who had worked on Google's then-new in-house routing system for Google Maps (the system that generates directions between two locations). Before, they licensed a routing system from a third party, and it was expensive ($) and slow.
The in-house system was cheap enough to be almost free in comparison, and it produced results in tens of milliseconds instead of many hundreds or even thousands of milliseconds. That allowed Google to build the amazing-at-the-time "drag to change the route" feature that would live-update the route to pass through the point under your cursor. It ran a new routing query many times per second.
Large Language Model + Large Earth Model
Shameless plug: recently we've built a demo that allows you to search for objects in San Francisco using natural language. You can look for things like Tesla cars, dry patches, boats, and more. Link: https://demo.bluesight.ai/
We've tried using Clay embeddings but we quickly found out that they perform poorly for similarity search compared to embeddings produced by CLIP fine tuned on OSM captions (SkyScript).
We did try to relate OSM tags to Clay embeddings, but it didn't scale well. We did not give up, but we are re-considering ( https://github.com/Clay-foundation/earth-text ). I think SatClip plus OSM is a better approach. or LLM embeddings mapped to Clay embeddings...
We tried to search for bridges, beaches, tennis courts, etc. It worked, but it didn't work well. The top of the ranking was filled with unrelated objects. We found that similarity scores are stacked together too much (similarity values are between 0.91 and 0.92 with 4 digit difference, ~200k tiles), so the encoder made very little difference between objects.
I believe that Clay can be used with additional fine-tuning for classification and segmentation, but standalone embeddings are pretty poor.
Check this: https://github.com/wangzhecheng/SkyScript. It is a dataset of OSM tags and satellite images. CLIP fine-tuned on that gives good embeddings for text-to-image search as well as image-to-image.
…we’ve built our own foundation model, GFT (Generative Forecasting Transformer), a 1.5B parameter frontier model that simulates global weather…
Hence: simulate the Earth.
Best of luck, and thanks for taking the leap! Humanity will surely thank you. Hopefully one day you can claim a bit of the NWS’ $1.2B annual budget, or the US Navy’s $infinity budget — if you haven’t, definitely reach out to NRL and see if they’ll buy what you’re selling!
Oh and C) reach out if you ever find the need to contract out a naive, cheap, and annoyingly-optimistic full stack engineer/philosopher ;)
Re question 2: Simulations don't need to be explainable. Being able to simulate simply means being able to provide a resonable evolution of a system given some potential set of initial conditions and other constraints. Even for physics-based simulations, when run at huge scale like with weather, it's debatable to what degree they are "interpretable".
Thanks for your questions!
[1] https://x.com/karpathy/status/1835024197506187617 [2] https://www.youtube.com/watch?v=-KMdo9AWJaQ&t=1010s
What exactly is predicted and what is the actual path in those videos?
For one, I always thought it would be informative for things like game engines to have a reference point. How fast to streams typically flow in this type of environment? What tree species are even in this geo?
https://data.neonscience.org/data-api/graphql/explorer/build...
Where we're going, we don't need "Data Products".
I had a web app online in 2020-22 called Skim Day that predicted skimboarding conditions on California beaches that was mostly powered by weather APIs. The tide predictions were solid, but the weather itself was almost never right, especially wind speed. Additionally there were some missing metrics like slope of beach which changes significantly throughout the year and is very important for skimboarding.
Basically, I needed AI. And this looks incredible. Love your website and even the name and concept of "Generative Forecasting Transformer (GFT)" - very cool. I imagine the likes of Surfline, The Weather Channel, and NOAA would be interested to say the least.
What will your differentiators be?
Are you paying for weather data products?
Better weather predictions are worth money, plain and simple.
Once upon a time I converted spectral-transform-shallow-water-model (STSWM or parallelized as PSTSWM) from FORTRAN to Verilog. I believe this is the spectral-transform method we have run for the last 30 years to do forecasting. The forecasting would be ~20% different results for 10-day predictions if we truncated each operation to FP64 instead of Intel's FP80.
1. The truth is we still have to investigate the the numerical stability of these models. Our GFT forecast rollouts are around 2 weeks (~60 steps) long and things are stable in in that range. We're working on longer-ranged forecasts internally.
2. The compute requirements are extremely favorable for ML methods. Our training costs are significantly cheaper than the fixed costs of the supercomputers that government agencies require and each forecast can be generated on 1 GPU over a few minutes instead of 1 supercomputer over a few hours.
3. There's a similar floating-point story in deep learning models with FP32, FP16, BF16 (and even lower these days)! An exciting area to explore
1. How will you handle one-off events like volcanic eruptions for instance? 2. Where do you start with this too? Do you pitch a meteorology team? Is it like a "compare and see for yourself"?
Re where do we start. A lot of organisations across different sectors need better weather predictions or simulations that depend on weather. Measuring the skill of such models is a relatively standard procedure and people can check the numbers.
Haha. The old NLP saying "every time I fire a linguist, my performance goes up", now applies to the physicists....
Disclosure: I work there.
Have specific industries reached out to you for your commerical potential – natural resource exploration, for example?
As a fellow deep learning modeler of Earth systems, I can also say that what they're doing really is 100% top notch. Congrats to the team and YC.
Using the full expressive power of a programming language to model the real world and then execute AI algorithms on highly structured and highly understood data seems like the right way to go!