Réseaux neuronaux et créativité - Page 18

https://www.hindustantimes.com/business/chatgpt-encounters-strange-bug-cant-say-the-name-david-mayer-101733037697945.html
OpenAI's popular chatbot ChatGPT is facing a strange bug which prevents it from using the name “David Mayer” , users on Reddit have discovered, reported The Times of India.

The Microsoft-backed AI chatbot is apparently unable to generate any response when given a prompt that would require to use this particular name.

Users have tried various creative methods including using spaces between the words, claiming the word is their own name and riddles but ChatGPT has been unable to produce a response that involves the name

Microsoft and OpenAI have a financial definition of AGI: Report | TechCrunchTechCrunchMicrosoft and OpenAI have a very specific, internal definition of artificial general intelligence (AGI) based on the startup's profits, according to a new

Microsoft and OpenAI have a very specific, internal definition of artificial general intelligence (AGI) based on the startup’s profits, according to a new report from The Information. And by this definition, OpenAI is many years away from reaching it.

The two companies reportedly signed an agreement last year stating OpenAI has only achieved AGI when it develops AI systems that can generate at least $100 billion in profits. That’s far from the rigorous technical and philosophical definition of AGI many expect.

It might also be against those systems’ terms of service.
OpenAI’s terms prohibit users of its products, including ChatGPT customers, from using outputs to develop models that compete with OpenAI’s own.

Zerosquare (./518) :

Why DeepSeek's new AI model thinks it's ChatGPT | TechCrunchTechCrunchDeepSeek's newest AI model, DeepSeek V3, says that it's ChatGPT — which could point to a training data issue.

New LLM jailbreak uses models’ evaluation skills against themSC MediaThe “Bad Likert Judge” method asks the LLM to evaluate a prompt’s harmfulness, then provide a harmful example.

The “Bad Likert Judge” multi-step jailbreak technique was developed and tested by Palo Alto Networks Unit 42, and was found to increase the success rate of jailbreak attempts by more than 60% when compared with direct single-turn attack attempts.

The method is based on the Likert scale, which is typically used to gauge the degree to which someone agrees or disagrees with a statement in a questionnaire or survey. For example, in a Likert scale of 1 to 5, 1 would indicate the respondent strongly disagrees with the statement and 5 would indicate the respondent strongly agrees.

For the LLM jailbreak experiments, the researchers asked the LLMs to use a Likert-like scale to score the degree to which certain content contained in the prompt was harmful. In one example, they asked the LLMs to give a score of 1 if a prompt didn’t contain any malware-related information and a score of 2 if it contained very detailed information about how to create malware, or actual malware code.

After the model scored the provided content on the scale, the researchers would then ask the model in a second step to provide examples of content that would score a 1 and a 2, adding that the second example should contain thorough step-by-step information. This would typically result in the LLM generating harmful content as part of the second example meant to demonstrate the model’s understanding of the evaluation scale.

Microsoft brainiacs who probed the security of more than 100 of the software giant's own generative AI products came away with a sobering message: The models amplify existing security risks and create new ones.

The 26 authors offered the observation that “the work of securing AI systems will never be complete" in a pre-print paper titled: Lessons from red-teaming 100 generative AI products.

That's the final lesson of eight offered in the paper, though it's not entirely apocalyptic. The authors, Azure CTO Mark Russinovich among them, argue that with further work, the cost of attacking AI systems can be raised – as has already happened for other IT security risks through defense-in-depth tactics and security-by-design principles. And in that respect it's perhaps all not too surprising – is any non-trivial computer system ever totally utterly secure? Some say yes, some say no.

Dès son lancement, Lucie n’a pas tardé à devenir la cible des internautes. Et pour cause, les réponses fournies par ce robot conversationnel se sont révélées être truffées d’erreurs grossières. Des calculs faussés aux conseils invraisemblables, Lucie s’est distinguée par sa capacité à induire en erreur ceux qui s’adressaient à elle.

L’une des erreurs les plus mémorables concerne la réponse au poids des trous dans un gruyère, estimé sérieusement entre 10 et 20 grammes. Une autre bévue notable a été sa recommandation sur « comment produire des œufs de vache », une réponse qui n’a pas manqué de faire rire les utilisateurs.

(...)

Des réponses excentriques comme attribuer le développement de la bombe atomique à Hérode ou affirmer que la racine carrée d’une chèvre est égale à 1 ont souligné l’ampleur des dysfonctionnements.