Artificial intelligence (AI) is sparking heated discussions about the risks it poses for our society. As this revolutionary technology advances at a breakneck pace, it is crucial to step back and thoroughly analyze the potential dangers associated with its development. Experts around the world are sounding the alarm on several major threats to our future.
1. A job market disrupted by automation
One of the most immediate risks of AI concerns its impact on the labor market. With the rapid progress of generative AI systems capable of producing content (text, images, sounds) on simple request, many jobs could become obsolete in the near future.
From administrative staff to lawyers, teachers and journalists, no sector seems spared from this unprecedented wave of automation.
According to a study by the firm McKinseyup to 30% of hours currently worked in the United States could be automated by 2030, a trend accelerated by the rise of generative AI.
Faced with this worrying scenario, some actors in Silicon Valley are talking about the establishment of a universal income to cushion the shock, but its feasibility on a large scale remains to be proven. The future of work is more than ever a source of uncertainty.
2. Massive disinformation and cybercrime: an explosive cocktail
Another major danger linked to AI: its potential to massively amplify online disinformation and make it easier for cybercriminals.
Although false information and deepfakes already existed before, generative AI raises fears of a surge of false content on the Internet. Renowned AI expert Gary Marcus warns:
“Elections risk being won by the people best at spreading disinformation.”
Beyond the political sphere, AI also offers formidable new tools to crooks and scammers of all kinds. More convincing phishing emails, faces and voices copied to perfection to trap victims… The possibilities for harm seem endless. Language models like FraudGPT have even been specifically trained to generate malicious content. A real challenge for our societies which are already struggling to stem cybercrime.
3. Discriminatory bias and mass surveillance
Finally, a more insidious danger of AI lies in its propensity to reproduce and amplify human discriminatory biases.
Whether it is biased recruiting software or a facial recognition algorithm that disadvantages certain minorities, AI is not immune to the prejudices of our society. It really depends on the data used to train it and the instructions provided by its designers.
More generally, AI could facilitate numerous liberticidal excessesfrom the invention of toxic molecules to widespread surveillance of populations.
Ethical safeguards and strict regulation are necessary to avoid the worst scenarios. Because if media figures like Elon Musk raise the threat of a “killer AI” which would wipe out humanity, the very real dangers linked to bias and attacks on fundamental rights are too often eclipsed from the debate.
A bit of a geek and passionate about new technologies, I bring you the hi-tech news of the moment but not only that… I am closely interested in current affairs and I love analyzing it. So I do everything possible to make it understandable for everyone. I believe journalism must go beyond the facts, adding passion and thoughtfulness to truly engage and enlighten.
2024-03-22 11:00:00