In partnership with
News of the day
1. Anthropic's new Channels feature for Claude Code enables autonomous AI agent operation by integrating external events from Telegram, Discord, or custom MCP channels directly into active sessions, allowing Claude to monitor and respond to CI results, alerts, and messages even when no one is at the terminal. → Read more
2. Mistral AI CEO Arthur Mensch proposes a mandatory revenue-based contribution from European AI providers to fund cultural creation and ensure legal certainty for AI training. → Read more
3. OpenAI aims to build an autonomous AI researcher capable of solving complex scientific problems, with prototypes by September and 2028, raising ethical concerns. → Read more
4. Palantir's developer conference highlighted AI's role in warfare, with CEO Alex Karp emphasizing an "unfair advantage" for troops. Generative AI fuels growth in defense and commercial sectors, with a focus on enabling victory. → Read more
Our take
Hi Dotikers!
Yesterday, we were questioning Apple's positioning in a world saturated with cloud inference capabilities, and what hardware-software integration says about the real value of a platform. Today, Anthropic is asking a similar question, but from the developer tools side: should an AI agent wait to be spoken to, or can it just work?
With "Channels," Claude Code shifts from a terminal assistant to an always-on agent. The concept is surgical: an MCP server connects a Claude Code session running in the background to an external messaging platform, Telegram or Discord for now. The agent receives a message, executes the task in the user's local environment (files, git, tools) and responds in the same channel. No dedicated web interface, no intermediary cloud. Your machine does the work, and you get the result on your phone while you grab a coffee.
This is a direct response to OpenClaw. Peter Steinberger's open-source project, which had won over thousands of developers precisely because it let them control an agent from WhatsApp or Telegram without going through a proprietary interface, now finds itself challenged on its own turf. Some early adopters on X said it plainly: no more need to buy a dedicated Mac Mini to run an agent 24/7. Anthropic has baked in the same promise, with the robustness of a tier-one service as a bonus.
The comparison stops there, however. OpenClaw remains open, hackable, freely deployable. Channels require a Claude account, don't support API keys, and enterprise teams must manually enable the feature through admin settings. Anthropic has built something polished, well-documented, and opinionated. It's a commercial strength, not an open-source philosophy.
What's taking shape is a gradual normalization of the asynchronous agent in the developer workflow. AI no longer just responds. It runs. And you interact whenever it suits you.
Alex.
Here’s how I use Attio to run my day.
Attio is the AI CRM with conversational AI built directly into your workspace. Every morning, Ask Attio handles my prep:
Surfaces insights from calls and conversations across my entire CRM
Update records and create tasks without manual entry
Answers questions about deals, accounts, and customer signals that used to take hours to find
All in seconds. No searching, no switching tabs, no manual updates.
Ready to scale faster?
Meme of the day
