Yes, AI Models Can Get Worse over Time

3 août 2023

More training and more data can have unintended consequences for machine-learning models such as GPT-4

When OpenAI released its latest text-generating artificial intelligence, the large language model GPT-4, in March, it was very good at identifying prime numbers. When the AI was given a series of 500 such numbers and asked whether they were primes, it correctly labeled them 97.6 percent of the time. But a few months later, in June, the same test yielded very different results. GPT-4 only correctly labeled 2.4 percent of the prime numbers AI researchers prompted it with—a complete reversal in apparent accuracy. The finding underscores the complexity of large artificial intelligence models: instead of AI uniformly improving at every task on a straight trajectory, the reality is much more like a winding road full of speed bumps and detours.
The drastic shift in GPT-4’s performance was highlighted in a buzzy preprint study released last month by three computer scientists: two at Stanford University and one at the University of California, Berkeley. The researchers ran tests on both GPT-4 and its predecessor, GPT-3.5, in March and June. They found lots of differences between the two AI models—and also across each one’s output over time. The changes that just a few months seemed to make in GPT-4’s behavior were particularly striking.

Mots-clés : cybersécurité, sécurité informatique, protection des données, menaces cybernétiques, veille cyber, analyse de vulnérabilités, sécurité des réseaux, cyberattaques, conformité RGPD, NIS2, DORA, PCIDSS, DEVSECOPS, eSANTE, intelligence artificielle, IA en cybersécurité, apprentissage automatique, deep learning, algorithmes de sécurité, détection des anomalies, systèmes intelligents, automatisation de la sécurité, IA pour la prévention des cyberattaques.