Deepmind’s Newest AI doing the Impossible! – Perfect Weather Predictions

Deep mind is pushing its industry leading artificial intelligence into a new difficult field in the hopes of surpassing the tens of thousands of people who are working in that field. After conquering go in starcraft now, deepmind has demonstrated their newest ai model, which is expected to be one of the most revolutionary inventions in the past 20 years. So what exactly can it do? It has a 99 99 accuracy in predicting the weather in

Any location this is a huge step up from the 74 accuracy which most people working in the field managed to hit, and this is not even all it can do. It can even teach us how to influence the weather, in certain conditions. Welcome to todays episode of ai news in this episode, i will show you the most recent artificial intelligence model and its exact capabilities from googles deep mind what else this can do for the meteorology field and, finally, how society could change in the event of

Artificial intelligence completely taking over a job without humans being able to compete in any way and whether or not as having the ability to change the weather is a good thing or a bad thing. The alphabet owned business has been discreetly collaborating with the met office for a few years, and the results of their partnership were published today in the journal nature. In a nutshell, deepmind has developed a new machine learning model that can

Forecast, whether or not it will rain in the following future precipitation now casting is a kind of weather forecasting that predicts rainfall on a very short period up to two hours before a storm. Todays weather forecasts are quite good at predicting rain in the future. Anything from six hours to a couple weeks, however, any sooner than that blind spots develop and here is where machine learning may help fill in the gaps. A downpour isnt only inconvenient for someone who recently had their hair done.

the ability to predict heavy rain in advance is critical for daily yet critical circumstances including road safety air traffic control and flood early warning systems extreme weather occurrences such as severe rain storms or flooding will only grow more common as the climate crisis worsens the capacity to anticipate rainfall more accurately and quickly is critical for making prompt decisions in certain scenarios such as stopping a train or evacuating a building the data was put into a deep generative network dgn which is a type.

Of ai that is trained to produce new data samples that are remarkably close to the original data it was trained on imitation faces, including fake rembrandts, have been created using gans dgmr, which stands for deep generative model of rainfall learned to create false radar. Pictures that followed the sequence of actual observations in this example, according to shakir muhammad, who lead the deep mind study its similar to viewing a few frames of a movie and

Anticipating what would happen next according to nyle robinson head of partnerships, and, product innovation at the met office who co-authored the paper, machine learning algorithms usually aim to optimize for one basic metric of how excellent their forecast is however weather forecasts may be beneficial or harmful in? A variety of ways, perhaps one prediction – predicts precipitation in the appropriate spot, but at the wrong intensity, while another

Predicts the proper mix of intensities but in the wrong regions and so. On we spent a lot of time and effort in this. Study evaluating our algorithm against a variety of metrics radar, imagery is used by the met office to forecast when the skies will open up radar works by firing a beam into the atmosphere and measuring the time it

Takes for it to return which informs you, how much moisture is there? There will be more rain if there is more moisture. The data is then transferred to the met offices headquarters where it is analyzed to provide a picture of precipitation across the united kingdom. Deepminds model was trained using radar images from the united kingdom between 2016 and 2018, allowing it to anticipate

what would happen in an hour or two the researchers also tested their model with more than 50 met office meteorologists the meteorologists were instructed to compare deepminds dgmr technique against another now casting method called as pi steps as well as another deep learning method in a blind research in over 90 of situations deepminds model outperformed the others in terms of accuracy and utility according to the researchers academic article existing supercomputers are accurate at forecasting low intensity rainfall using data such as wind speeds the new technique which relies on computers to understand connections between radar data and rainfall performs.

better in medium to heavy rain events which are becoming more common as a result of climate change well be able to forecast how unpredictable the world has become as a result of our actions deepminds projections are only accurate up to 90 minutes ahead of time and the company hasnt revealed how much better they are than current techniques even still as soon as ai does anything its reasonable to wonder if it will be.

Able to accomplish everything, will we one day have totally accurate weather forecasts? Would we want to live in a world if we knew the weather every day of our lives? Is that how it is in australia, because weather systems are so complicated its not going to happen? However, the death of uncertainty would be odd yeah. We have wedding availability in june 2024, but dont book, since its likely to rain for better or worse wed, be less

startled because everything would be properly organized there would be fewer cancellations and our lives would become even busy if there was no uncertainty would we talk about the weather less or more more importantly i believe that those who have read the prognosis will wish to share the message the likelihood of rain would remain the apolitical icebreaker it is now of course its strange that we enjoy discussing the weather yet hate discussing the environment would i rather know the weather for the.

next two hours or influence the environment for the next two decades a prediction is only as good as the response it generates other scientists arent as taken aback by the findings a meteorologist from the university of redding said i dont see any revolution in predicting here and he is perplexed by their reliance on this statistic to demonstrate the utility of their approach hes still perplexed as to why they didnt utilize a more suitable objective score there isnt much information on how the assessments were conducted or even what was assessed deep mind does not provide a specific.

figure for how much more accurate their model is compared to other models rather than reporting a single figure we wanted to adopt a more measured approach according to deep mind deep mind study might just give an alternative method to rain forecasting rather than entirely disrupting the industry as we know it it performs similarly to other cutting-edge versions now on the market however their model isnt much ahead in fact it may be somewhat superior a meteorologist from the university of manchester david schultz agrees the suggested method appears to outperform them is this a game changer.

im not sure is this a significant improvement over some of the current techniques in their paper yes several newspapers and magazines including the uks sunday times and arabian business reported in january 2011 that scientists backed by the government of abu dhabi the capital of the united arab emirates had created over 50 artificial rainstorms near al ain the countrys second largest city between july and august 2010 near the countrys border with oman the man-made rainstorms were claimed to have occasionally produced hail gales and thunderstorms perplexing locals.

cloud seeding as a popular method for increasing precipitation cloud seeding is the practice of spraying tiny particles onto clouds such as silver iodide in order to influence their formation generally with the objective of enhancing precipitation although there are no immediate plans to utilize the model the team aims to use it to influence future weather reports in the future however the researchers claim that ai would never be able to completely replace weather forecasts youll need specialists and humans in.

the loop to ensure that what theyre seeing in terms of forecasts is reasonable and then how to convey it to the general public its difficult to see an automated system performing all of these tasks with the required delicacy and competence deepminds partnership with the med office is an excellent example of ai development that involves input from the end user which seems like a no-brainer but seldom happens the team worked on the project for several years with professional advice from the med office shaping it it drove our model development in a different direction than we would have.

gone on our own says deepmind researcher suman rivori otherwise we may have created a model that was eventually useless deep mind is likewise keen to show that its artificial intelligence has real world applications dgmr like alpha fold is part of shakirs story the firm is cashing in on its years of solving difficult issues in gaming the most important message here is that deep mind is now starting to cross real.

world science issues off its bucket list so what is your opinion on deepminds newest and most advanced ai model do you believe it can take over an entire industry do you believe its ability to teach us how to change the weather and if its possible in the first place to accurately predict the weather over a long period of time should we be able to do that in the first place please share your opinion with us in the comment section below i.

would love to hear what you have to say about it thank you for watching ai news we consistently report on the newest technologies that are shaping the future of our world wed appreciate you subscribing and watching our other videos see you around and take care.