TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Google says AI weather model masters 15-day forecast

286 pointsby lemonberry5 months ago

34 comments

Ankaios5 months ago
Turns out the Google AI&#x27;s 15-day forecasts are for the weather in Mountain View.<p>Which I can forecast 15 days out, too.
评论 #42385798 未加载
评论 #42386755 未加载
评论 #42388903 未加载
评论 #42385014 未加载
评论 #42385549 未加载
评论 #42388734 未加载
评论 #42387291 未加载
评论 #42387214 未加载
评论 #42385397 未加载
atonse5 months ago
This is great from a practical standpoint (being able to predict weather), but does it actually improve our understanding of the weather, or WHY those predictions are better?<p>That is my issue with some of these AI advances. With these, we won&#x27;t have actually gotten better at understanding the weather patterns, since it&#x27;s all just a bunch of weights which nobody really understands.
评论 #42384077 未加载
评论 #42382504 未加载
评论 #42384201 未加载
评论 #42381418 未加载
评论 #42382409 未加载
评论 #42381554 未加载
评论 #42381948 未加载
评论 #42382324 未加载
评论 #42381824 未加载
评论 #42381516 未加载
评论 #42383731 未加载
评论 #42387044 未加载
评论 #42381448 未加载
评论 #42385835 未加载
评论 #42383782 未加载
评论 #42381431 未加载
评论 #42383765 未加载
评论 #42383990 未加载
评论 #42385470 未加载
评论 #42381515 未加载
评论 #42384310 未加载
BadHumans5 months ago
I&#x27;m friends with a meteorologist and the 15+ day forecast is the bane of their existence because you can&#x27;t accurately forecast beyond a week so I would love to know how they are measuring accuracy. The article doesn&#x27;t say and I know the paper is going to go over my head.
评论 #42386583 未加载
评论 #42382245 未加载
评论 #42386774 未加载
评论 #42387076 未加载
评论 #42382838 未加载
Always425 months ago
Based on the Gemini release it’s hard to take what Google claims at face value.
评论 #42382536 未加载
评论 #42381944 未加载
评论 #42382021 未加载
评论 #42381580 未加载
评论 #42381763 未加载
评论 #42384529 未加载
评论 #42381587 未加载
queuebert5 months ago
They can say what they want, but I get rained on by surprise rain more than I ever have in my life, now that I&#x27;m practically forced to use the built-in Google weather due to them and Apple catching and killing all the good weather apps.
评论 #42384182 未加载
评论 #42385343 未加载
评论 #42387496 未加载
xnx5 months ago
6 days ago: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=42319062">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=42319062</a>
tokai5 months ago
Will never trust Googles claims about forecasting since the google flu debacle. They&#x27;ll fudge the numbers until they suit them.
choeger5 months ago
Is that a prediction in the form of &quot;There&#x27;s a 30% chance of light rain that day.&quot; or &quot;Temperature will reach 22,5°C at 14:00.&quot;?
评论 #42385587 未加载
ericra5 months ago
It does seem like this is one of those domains where new AI models could thrive. From my understanding, the amount and variety of data necessary to make these models work is huge. In addition to historical data, you&#x27;ve got constant satellite data, weather stations on the ground for data collection, weather balloons going high into the atmosphere multiple times daily per location, Doppler radars tracking precipitation, data from ships and other devices in the ocean measuring temps and other info, and who knows what else.<p>It&#x27;s incredible that we are able to predict anything this far into the future, but the complexity seems like it lends itself to this kind of black box approach.<p>*This is all speculation, so I&#x27;d be grateful if someone more knowledgeable could tell if if I&#x27;m mistaken about these assumptions. It&#x27;s an interesting topic.
评论 #42384534 未加载
YeGoblynQueenne5 months ago
The catch: the model is predicting the past:<p>&gt;&gt; But DeepMind said GenCast surpassed the precision of the center&#x27;s forecasts in more than 97 percent of the 1,320 real-world scenarios from 2019 which they were both tested on.
评论 #42386084 未加载
amelius5 months ago
Can&#x27;t we let prediction markets solve the weather forecasting problem?
mnky9800n5 months ago
I don’t really understand why Google and other companies making similar models are able to train on existing modelled or reanalysis data sets and then claim further accuracy than the originals. Sure, stacks of convolutions with multimodal attention blocks should be able to tease apart all the of idiosyncratic correlations that the original models may not have seen. But it’s unclear to me that better models is the direction to go in as opposed to better data.
评论 #42388692 未加载
mattxxx5 months ago
Can we extract the latent understanding it has?<p>Would be really cool to convert it&#x27;s predictive model into a computer program that predicts written in like python&#x2F;C&#x2F;rust&#x2F;whatever, and I think that would better serve our ability to understand the world.
评论 #42384376 未加载
helsinki5 months ago
On a related note, the company I work for is hiring a meteorologist for a commodities trading desk in NYC and London. Send me an e-mail if you would like a referral: <a href="https:&#x2F;&#x2F;mlp.eightfold.ai&#x2F;careers&#x2F;job?domain=mlp.com&amp;pid=755939667061&amp;domain=mlp.com&amp;sort_by=relevance&amp;job_index=84" rel="nofollow">https:&#x2F;&#x2F;mlp.eightfold.ai&#x2F;careers&#x2F;job?domain=mlp.com&amp;pid=7559...</a>
smartmic5 months ago
Discussed 3 days ago: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=42351698">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=42351698</a>
dumbfounder5 months ago
&quot;More accurate forecasts of risks of extreme weather can help officials safeguard more lives, avert damage, and save money,&quot; DeepMind said.<p>Did the DeepMind AI say this?
评论 #42381881 未加载
评论 #42383544 未加载
wslh5 months ago
I really wonder if, like the weather, we should see new financial prediction models, understanding that there is randomness but also patterns.
评论 #42388547 未加载
评论 #42385542 未加载
encoderer5 months ago
I recently read that the 10 day forecast now is as accurate as the 5 day forecast from 30 years ago when I was a kid.<p>This surprised me. I grew up in Ohio and now I live in the Bay Area and the forecast here seems to be accurate only 2-3 days out. It would be so helpful to have an accurate 10 day forecast.
评论 #42383457 未加载
评论 #42383050 未加载
BrawnyBadger535 months ago
I thought I remembered this model being released many months ago? Or is this a new update to the model?
评论 #42381107 未加载
评论 #42381559 未加载
Aeroi5 months ago
Sweet, how will this help for commodities futures trading?
wiz21c5 months ago
FTA:<p>&gt; But he said these forecasting systems are reliant on the weather prediction models that are already running, such as that operated by ECMWF.<p>et voilà.
bilsbie5 months ago
Could they train this as a multi model model with language so you could actually ask it questions about its predictions and such?
space_oddity5 months ago
How long before models like this are reliably part of decision-making for disaster response or everyday weather planning?
motoboi5 months ago
This is from a model from 2019! Published now on nature. Imagine what they are cooking now.
Aeroi5 months ago
Does anyone know if this works for marine forecasts? specifically weather gribs?
fancyfredbot5 months ago
It takes 8 minutes to produce a 15 day forecast. That&#x27;s actually quite a long time for an AI model. I should probably read the paper to find out why but does anyone know? Is the model predicting the weather in 10 minutes time and just run iteratively 2000 times for a 14 day forecast?
评论 #42384553 未加载
gwbas1c5 months ago
This makes me wonder: How far back can weather be reliably predicted?
评论 #42389103 未加载
评论 #42388735 未加载
urbandw311er5 months ago
As long as the AI in charge of weather forecasting never ever comes within the control of the AI charged with fixing climate change.<p>GOAL: Fix world’s problems. OBSERVED CAUSE: Too many humans. DECISION: Tell humans about impending hurricane? Yes&#x2F;No
评论 #42385980 未加载
bad_haircut725 months ago
Obviously this team knows way more about this donain than me but I have to ask, wouldnt this only be able to predict weather which is in line with past weather patterns&#x2F;indicators? I can imagine a weather analyst might be able to see &quot;between the data&quot; and recognise when some anomaly might be brewing, but an AI model would not
评论 #42381352 未加载
评论 #42381648 未加载
评论 #42381470 未加载
评论 #42384104 未加载
butlike5 months ago
Why not 16?
jdlyga5 months ago
Yeah, but can their AI weather model show up in a cute outfit and point to a map too
评论 #42394477 未加载
ovj1295 months ago
This a great use case for AI to be used
brcmthrowaway5 months ago
Meteorologists are quaking in their boots!
评论 #42384140 未加载
latexr5 months ago
&gt; A new artificial intelligence-based weather model can deliver 15-day forecasts with unrivaled accuracy and speed, a Google lab said, with potentially life-saving applications as climate change ramps up.<p>How hilariously dystopian. Instead of spending energy&#x2F;power&#x2F;influence to mitigate&#x2F;solve&#x2F;revert the problem, they’re actively making it worse to in the name of predicting how bad it’s going to be (to try and save themselves, which will ultimately fail).
评论 #42387200 未加载
评论 #42386995 未加载