
Tech • AI • Robotics • Game
A French tech commentator used a live session to flag major changes at Anthropic’s Claude, promote open local AI setups, and announce a public France Télévisions event on AI scheduled for 22 October 2026 in Paris.
Anthropic is rolling out a merger between Chat and Claude Code or Cowork-style workflows, removing the manual toggle that currently separates simple conversation from project-based file work. The change is presented as a usability improvement, with the model expected to route requests intelligently on its own. The concern is that this could deepen the black-box effect, making it harder for users to know when the system is searching project files, when it is answering from general knowledge, and how to force one behavior over the other.
A second interface change introduces new output options including Doc, Slide, Design, and Artifact in beta. These features move Claude closer to office productivity and lightweight design generation, letting users create structured documents and presentations directly in the interface. The practical limitation is that these outputs remain tied to Claude’s own environment rather than automatically becoming standard local files, which may complicate workflows for users expecting direct export and full control.
The discussion highlighted growing interest in open-source, local AI systems that keep data on the user’s own machine rather than in external cloud services. The trade-off is clear: strong local inference requires open-weight models and a powerful computer with enough memory and processing capacity, making entry costs high. Even so, local deployment was presented as a credible alternative to dependence on subscription-based services from major US platforms.
A recent corporate conference in Paris was described as a sign that AI use inside companies is entering a more concrete phase. Attendees reportedly included both skeptical participants and more advanced practitioners, but the shared focus was no longer whether AI mattered. The key question had become how organizations should deploy it internally, with an emphasis on collective experimentation rather than top-down imposition.
The session clarified that opposition to AI is not being dismissed, but that criticism should be fact-based, coherent, and technically informed. The distinction drawn was between principled critiques and broad claims that collapse under basic verification. That framing reflects a wider debate in France and elsewhere, where concerns about AI’s social and economic effects are growing but often compete with exaggerated or weakly sourced arguments.
Among the most serious issues identified were data centers, energy consumption, and the broader ecological footprint of AI systems. These concerns were treated as legitimate and urgent rather than marginal. To address them, an upcoming interview was announced with The Shift Project, a think tank known for work on energy transition and climate constraints, signaling an effort to ground the AI debate in measurable environmental realities.
A major announcement concerned participation in a France Télévisions program on AI in overseas territories, to be recorded in Paris on 22 October 2026 from 13:30 to 18:00. The event will include several contributors, field reports, and discussions on themes such as education, employment, and entrepreneurship. Audience seats will be available free of charge, though numbers are limited.
On the morning of 22 October 2026, a separate live segment is planned from inside France Télévisions, with a behind-the-scenes look at studios, teams, and production spaces. The aim is to show how the broadcaster operates while linking that visit to the afternoon’s AI-focused recording. The initiative also reflects a broader institutional push by France Télévisions to engage seriously with AI rather than treating it as a passing headline topic.
The session underscored two parallel trends: AI tools are becoming more powerful but also more opaque, and public debate is shifting toward practical questions of governance, energy, and collective use. With Claude evolving rapidly and France Télévisions opening a national platform to these issues, the next phase of AI discussion appears set to be more public, more concrete, and more contested.
Ask a question