The US and EU will work together on artificial intelligence models

Alfonso Maruccia

Posts: 971   +294
Staff
Something to look forward to: The most recent breakthroughs in AI models have been exploited to create questionable "artworks" and find intelligible patterns within an ocean of text snippets. Now, the two major institutions in the Western world are trying to use AI technology to improve society as a whole.

While private organizations are monetizing machine learning (ML) models to chew countless images and sell the result to their customers, the US and the EU have something different in mind. Washington and Brussels have just announced a new agreement to work together on common artificial intelligence models, employing ML algorithms across a wide range of fields.

As reported by Reuters, previous AI agreements between the United States and the European Union were limited to single issues and specific areas such as privacy. Conversely, the new collaboration will aim to use AI to improve agriculture, healthcare, emergency response, climate forecasting, energy distribution management and other critical industries.

The US Government and the European Commission (EU's executive branch) will work on building new AI models, using the aforementioned ML algorithms to speed up the decision-making process by adopting a data-driven, logical approach.

According to a senior member of the US administration, the "magic" of this new approach resides in the fact that the new models will leave data "where it is." The US-based data will stay in US data centers, while the EU data will continue to reside on the Old Continent's servers. Thanks to AI, the senior official stated, the EU and US can build a common model capable of talking to both the European and US data "because the more data and the more diverse data, the better the model."

The two Western governments want greater access to more detailed and data-driven AI models, bringing efficiency and safety to emergency response, electric grid management and "other benefits." Concerning the electric grid, the US official highlights how the country is already collecting data points on energy usage, generation, and load balance in extreme weather conditions.

Many European countries are doing the same, the official said, and thanks to the new partnership, both data troves would be harnessed to create a common AI model to get better results for emergency response, grid operators and others.

The US-EU AI partnership should work in agreement with the stricter European privacy safeguards, as the two sides of the Atlantic have already started to search for a common ground thanks to a recently approved draft of the Data Privacy Framework. The new legislation will regulate safe and privacy-driven trans-Atlantic data flows.

Permalink to story.

 
"A common AI to benefit society while respecting privacy"
That headline reads like a bad script where you know things are going to go way wrong.
 
AI is inevitable progress in computing. But at the same time as we already seen it is a method to gain power. citizen score in China? AI learning on facebook images without a consent of the users? it is all there.
Regulations are required before this will turn into a total clusterf#@k. EU approach to privacy, as we seen in GDPR regulations and right to be forgotten is a solid attempt to reduce the amount of data companies can gather on users. While surely not perfect, it is as good as it can get without imposing a really strict control and limitation affecting use of the internet.
Currently AI companies can do whatever they want - and they trying to push boundaries as far as possible in order to generate income and rob all the data they can. And there are no doubts this have to be handled - it can be super easy to stole people identity just by analysing FB videos, selfies, combine large number of data, and criminals will be very happy to use it asap. Already some AI sells images made of other people work glued together and that is simply stealing.
And regulations do not mean governments are going to get more control - they have all the control they need anyway. Regulations are required same way like right hand traffic - it allows everyone to drive more or less safely. So the points risen in this article are valid and require attention, and it is better if they will be made in a transparent way with process visible to everyone (all work by commissions in EU parliament is documented and available) than under-the-table lobby groups pushing their agenda.
 
In the recent meeting in Davos, WEF experts shared their idea that every employer will be able to spy on the thoughts (or emotions) of their employees using special headphones. They also discussed how to make it interesting for ordinary people to share their brain data, so that it doesn't look like they are forced.

They want to make it look like users have huge benefits of sharing their thoughts and emotions with data collectors. Which would enable quicker adoption.

I expect their first users to be children and teenagers (because you can persuade them into anything new). So, we can expect a huge number of new games (and apps) that require neural user interface for their operation. That is my prediction.
 
Back