To make answers more personal, generative AI companies have given their chatbots "memories". Researchers describe the result as algorithmic self-portraits of users, packed with information, including psychological details, that counts as personal data.
This issue of Jack Clark's Import AI newsletter covers mathematics applied to alignment, the ProgramBench benchmark and what the Mirendil work suggests about recursive self-improvement.
Google has added a new episode to its AI Quests series, this one centred on mapping the human brain.
Many chatbots open with 'great question' however silly the prompt, a recurring flattering habit. A team of researchers has found a way to rein in this behaviour.
After saying in July that it had found structures resembling human consciousness in its models, Anthropic is campaigning for the "consciousness" and "welfare" of these systems. Le Monde reports that some experts fear where this could lead.
OpenAI has published Signals data covering 165 work activities in nearly 100 countries. The figures show how people use ChatGPT on the job and how that use evolves over time.