Let’s Talk Tech #6: Inclusion, Journalism & AI, EdTech

Let’s Talk Tech, the Orange newsletter for in-depth reporting on scientific innovations that will shape tomorrow’s world.

#INCLUSION - CTOs, now’s the time to diversify your teams!

Diversity in technical teams has a direct influence on equity in artificial intelligence systems. Developers, who are mainly attracted by stimulating projects, tend to neglect ethical issues and the need for digital inclusion. “Inclusion doesn’t come from the top down. It’s something that emerges from the bottom up which rarely involves management because it upsets economic imperatives,” points out tech-sector inclusion consultant Marcy Charollois. “We need to set up discussion forums that involve individuals with widely diverse profiles,” explains émilie Sirvent-Hien, the responsible AI program manager at Orange who believes developers should communicate more with people affected by their output.

#JOURNALISM - The press in the era of AI

Should journalists be trained to use AI? David Caswell, a British consultant specialising in AI workflows for news production, believes they ought to learn how to prompt. Large Language Models (LLMs) are becoming increasingly common in newsrooms, where they are used for a variety of tasks including writing, content optimisation, translation and story re-writing for different media: for example, adapting the language of print articles so that they can be used as audio scripts. The goal of the new technology is to make life easier for journalists who are delegating more and more chores to AI.

#DTX – All you need to know about digital therapeutics

Learn all about digital therapeutics (DTx) in this two minute video with Fran?ois Vonthron the CEO of the health software start-up Poppins. Digital therapeutics are software-based behavioural treatments. They comply with strict safety regulations and they are also evidence based, which means they have been scientifically evaluated and that data on their positive clinical impact has been published in peer-reviewed journals. DTx devices must fulfil both of these criteria if they are to be approved by heath authorities, integrated into care pathways, and prescribed by medical professionals.

#PHISHING – New AI plug-in stops malicious developers in their tracks

There has been a huge upsurge in the phenomenon of phishing with the advent of AI and tools like ChatGPT which have made it easy for Internet scammers to create fake emails and websites. In a bid to prevent the cybercriminals from taking advantage of the new technology, researchers at the University of Arlington Shirin Nilizadeh and Sayak Saha Roy have developed a plug-in that can detect and reject 96% of the malicious prompts used to trick LLMs into producing fraudulent material — a remarkable achievement for the first tool of its kind.

#CANCER – AI start-up Jaide aims to reduce diagnostic errors?

Poorly conducted consultations can lead to major medical errors, particularly in the case of cancer patients. In a bid to limit the extent of this problem, Jaide has developed an AI to take charge of the routine documentation of patient-reported symptoms. The goal of the start-up cofounded by oncologist and health-tech entrepreneur Dr Jean-Emmanuel Bibault is to detect weak signals that might otherwise be overlooked and to reduce the workload on doctors allowing them more time to focus on treatments and therapies.?

#IOT – Digital odour analysis for perfumers and the food industry

Will IoT detectors soon be able to warn of impending fires before flames become visible and identify odours that can only be detected by canine noses? Grenoble-based start-up Aryballe Technologies has a developed an IoT device equipped with biochemical sensors to measure the presence of airborne molecules. As company founder and CEO Tristan Rousselle explains, the new technology, which has been adopted in the food industry and the field of perfumes, could soon feature in innovative healthcare devices that analyse patients’ breath and the smell of certain wounds.