Affichez cet e-mail dans votre navigateur

The [artificial] intelligence newsletter

N°21

All forward-looking work begins with a deadline, because it's this deadline that pushes us to project, that forces us to build lines of continuity and envisage ruptures. Over the past fifty years, the pace of innovation has accelerated. That's why it's hard for technological thinking to project beyond 10-15 years. 2030 is therefore a conceivable timeframe.

The effervescence we've seen over the last five years in the field of artificial intelligence is due to the convergence of two phenomena: the digital revolution, whose corollary is the fact that (almost) every human activity involves a digital trace, and the 30,000-fold increase in computing power since 2020 ("Moore's Law").

We're all aware that the widespread use of this "new" technology (discovered in the 50s by Alan Turing) is already part of our daily lives. But what upheavals will it provoke? How will our societies come to terms with this technology, which claims to mimic -imperfectly- human thought?

The sudden irruption of AI into our lives raises the classic question: "Should we chain Prometheus?" And if we wanted to, could we?

Considering it illusory to believe that humanity could unlearn the knowledge it has acquired, we are condemned to tame the fire we have discovered. Scientific and technical discoveries are turning the world upside down and transforming our societies. We are convinced that mankind has an irreplaceable singularity, but we must collectively channel the uses of science to ensure that it remains at our service.

At Preligens, we are modest discoverers and users of this breakthrough technology. This obliges us to think beyond our daily practice, to define ethical lines and to contribute to collective reflection.

It's in this spirit that we're sharing the latest AI news, trends and discoveries in this new issue, so that you too can tell us what you expect from AI in 2030.

Enjoy your read!

The Preligens team

AI & satellite imagery: what about highly challenging detections

Artificial intelligence is a cutting-edge technology for automated processing of geospatial data. Applied to imagery intelligence, AI is a strategic capability for taking information superiority to the next level. But what happens when the images to be processed are complex? When the context, conditions or data are a real challenge for AI?

Discover the answer in our new video Tip&Cue #15 - 5 highly challenging detections performed by Preligens AI on our Youtube channel.

What’s new in AI?

World Trade Organization works on AI

As progress in AI shape new ways of working, many studies are looking at the jobs created or destroyed by new technologies. Recently, the World Trade Organization (WTO) examined the issue. Its finding is that "the first consequence of this new technology is unlikely to be job destruction, but rather potential changes in job quality, including work intensity and autonomy". AI thus offers new support for workers as they carry out their tasks. But some current jobs will undoubtedly be "automated", with AI replacing the employee. However, this phenomenon concerns only a small proportion of today's jobs (5.5%): mainly so-called "office jobs".

source

Military programming law: France adopts an AI budget for the Armed Forces

On July 13th 2023, the French Parliament adopted a new Military Programming Law for the years 2024 to 2030. Providing for an increase in the Armed Forces' budget and the acquisition of new equipment, the Law also creates a budget line dedicated to AI, to be debated each year when the Finance Acts are voted. This provision enables us to take into account the challenges of AI in military operations.

source

An international summit dedicated to AI  

On July 6 and 7, 2023, the United Nations organized a summit in Geneva dedicated to "AI for Social Good". The International Telecommunication Union (a UN agency) brought together more than 3,000 personalities (researchers, business leaders and humanoid robots) to discuss AI advances, opportunities and consequences. The United Nations called for global governance of AI to ensure the benefits of its development.

source

AI to diagnose cancer 

AI advances in medicine continue to develop. Some AI systems offer invaluable help in diagnosing certain diseases as early as possible. Such is the case with breast cancer, where early detection enables more effective treatment. An AI developed by ScreenPoint Medical can now automatically analyze mammograms to detect the presence of a tumor.  This technology reduces diagnosis time by around 45%, enabling faster, more effective treatment of patients affected by the disease.

source

When AI starts dubbing videos simultaneously

Artificial intelligence programs are multiplying with increasingly astonishing results. Two Tiktok veterans created HeyGen, a tool that automatically translates a video into several languages, respecting the intonation and movement of the speaker's lips. Ease of use which promises considerable time savings in terms of video editing, but which raises questions about the risk of ever more fake news appearing, making anyone say anything. Be wary of what you see on the internet… And what you hear!

source

3 questions to Cyril De Sousa Cardoso

AI in 2030: what could we expect?

What will AI look like in 2030? Some Preligens' employees imagine tomorrow's new AI applications for this short video. Discover their predictions and hopes in this video!  And you, what's your point of view about future of AI?

AI For Industry 2023

For the first time this year, AI for defense was one of the hot topics discussed on stage during the “AI for Industry” summit organized by Artefact on September 14.

Eve Doat-Arakelian - our VP Marketing & Public Affairs - shared the stage with Sébastien Massart - head of strategy at 3DS OUTSCALE - and Bruno RICAUD - AI Lead at Nexter - and addressed certain specific issues of AI applied to the sector of the defense.

Discover the replay of this round table on YouTube.