30초 요약
In June 2026, Matthew Berman reacts to Claude Tag, Anthropic's Slack-based AI agent, and predicts that agents such as Claude Code and Codex could move between existing applications and eventually reduce their interfaces to interchangeable back ends. His scenario that Anthropic and OpenAI could come to dominate software is a prediction, not an established outcome. The video explicitly says it is not an ad.
- 1Claude Tag in SlackBerman describes the new agent as working alongside a team in Slack; its actual scope depends on configuration and permissions.
- 2Agents as the interfaceHe expects Claude and Codex to appear throughout existing software and predicts people may interact primarily with agents.
- 3The database thesisBerman speculates that software interfaces could fade and agents might build their own databases, potentially concentrating power in AI companies.
핵심 정리
- Anthropic launched Claude Tag in Slack in June 2026, allowing teams to involve Claude in their work there.
- Berman predicts that AI agents will spread across existing software and increasingly become users' main interface.
- He speculates that agents could replace some applications with databases they create themselves.
- Claims that agents will outperform people across software or that two companies will own all software are forecasts, not demonstrated facts.
챕터별로 다듬은 전문 읽기
Claude Tag arrives
Berman introduces the Slack-based agent and describes its potential as a workplace assistant. Anthropic announced the product in June 2026.
Claude Tag has just launched, and I think it is a bigger deal than many people realize. Anthropic introduced it for Slack in June 2026, where teams can bring Claude into conversations and work with connected tools. I describe it as acting like an employee; that is an analogy, not a claim that the system has the responsibilities or capabilities of a human worker ([Anthropic's announcement](https://www.anthropic.com/news/introducing-claude-tag)).
Agents as the main interface
He predicts that agents will spread across software and users may increasingly interact with them instead of directly operating applications.
I expect Claude and probably Codex to appear in more of the software people already use. My prediction is that agents will increasingly operate those tools on the user's behalf, eventually making the agent the main interface. I also suggest they may do some work better than people using the applications directly, but this clip presents a forecast rather than a comparative test.
I think this points toward agents taking over much of knowledge work. In my scenario, the interfaces offered by software companies gradually matter less because users communicate directly with an agent. That is an extrapolation from a new product, not evidence that all knowledge work has already been automated.
A possible shift in software power
Berman imagines applications becoming back-end databases and agents creating alternatives. He predicts that the result could concentrate control in Anthropic and OpenAI.
Taking the idea further, I imagine applications becoming databases behind the agent. An agent might even offer to build a new database instead of relying on the existing provider. I speculate that this could allow Anthropic and OpenAI to dominate software; whether this happens depends on technical, commercial and governance factors the short video does not test.
같은 채널의 다른 글
34:21EN
Matthew Berman· 2026년 10월 27일Matthew Berman Examines Anthropic’s IPO Filing, AI Growth and GovernanceMatthew Berman, Sam Altman · Forward FutureThe presenter reviews the S-1’s financial metrics, compute strategy and voting structure while questioning the tension between rapid AI development and safety.
0:26ENMatthew Berman· 2026년 10월 25일Sam Altman on Human Values and Problem-Solving in an AI WorldSam Altman · OpenAIIn a short interview excerpt, Altman considers the enjoyment of solving hard problems and the human role in interpreting results.
0:36ENMatthew Berman· 2026년 10월 25일Kimi K3.1 and a Rubik’s Cube Solver DemonstrationMatthew Berman · Forward FutureMatthew Berman shows a scrambled virtual cube being solved and gives the Kimi K3.1 test a favorable assessment.
0:21ENMatthew Berman· 2026년 10월 24일An AI-Generated Kimi Benchmark Chart Used Invented ScoresMatthew Berman · Forward FutureMatthew Berman asks for a graph of Kimi K3.1 Pro benchmarks, then discovers that the displayed figures are illustrative placeholders.
25:50ENMatthew Berman· 2026년 10월 24일Kimi K3.1 Pro Tested: Strong 3D Demos, Uneven Results and Competitive PricingMatthew Berman · Forward FutureMatthew Berman tests the open-weight model on coding, 3D rendering and visual explanations, then compares benchmark results and prices.
32:20ENMatthew Berman· 2026년 10월 23일Who Is Local AI For? The Costs, Trade-offs and a Hybrid FutureMatthew Berman · Forward FutureMatthew Berman weighs data control and independence against hardware cost, setup complexity and inference speed.