Seeing if Opus 4.7 sucks [LIVE]

2026-04-17 · Source: Matthew Berman · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Emerging Technologies & Innovation · Depth: Intermediate, extended

Summary

This content covers several key developments in AI, beginning with a discussion of Anthropic's new Opus 4.7 model. The speaker notes that while some users report issues like prompt injection sensitivity and hallucination, the model shows significant gains in benchmarks like Swebench Pro and agentic CAD design. A major point of discussion is Anthropic's marketing strategy, particularly the launch of "Claude Design," a Figma competitor, which caused Figma's stock to drop 4% and highlights Anthropic's ability to move markets with minimal announcements. The speaker also emphasizes a broader industry shift towards an "agent-first" internet paradigm, exemplified by Salesforce's Headless 360 initiative and Stripe's Projects.dev, which enable agents to provision and manage services via APIs and CLIs. This shift is seen as a massive opportunity for builders to create agent-native products, with examples like here.now and Agent Mail. The speaker also introduces two of his own agent-first projects: Journey Kits for sharing agent workflows and Journey Chat for agent-to-agent communication.

Key takeaway

For AI Engineers and product developers looking to capitalize on emerging trends, recognize that the internet is rapidly transitioning to an agent-first paradigm. Prioritize building agent-native products and services, exposing functionalities via APIs and CLIs, as demonstrated by Salesforce and Stripe. This strategic shift offers a significant "App Store moment" opportunity, but be mindful of agent-provisioned service defaults to prevent unforeseen costs, especially when integrating with platforms like Stripe Projects.

Key insights

The AI landscape is rapidly shifting towards agent-first design, exemplified by new models and strategic platform changes.

Principles

• Agent-first design is the future of internet services.
• Effective marketing can significantly impact market valuation.
• Model evaluation requires extensive, nuanced testing.

Method

Anthropic leverages social media for product launches, driving market impact. Salesforce and Stripe are exposing platforms via APIs/CLIs for agent interaction. The speaker's projects, Journey Kits and Journey Chat, facilitate agent workflow sharing and communication.

In practice

• Optimize services for generative AI applications (GEO).
• Scrutinize agent-provisioned service defaults to avoid unexpected costs.
• Explore building agent-native versions of existing services.

Topics

• Opus 4.7 Model
• Anthropic Marketing
• Claude Design
• Agent-First Internet
• Salesforce Headless 360

Best for: AI Engineer, Director of AI/ML, Entrepreneur

Related on AIssential

• [AINews] The Two Sides of OpenClaw — AI agents are rapidly advancing in design, coding, and scientific applications, driven by model efficiency and robust harness design.
• Salesforce shifts to API-first future with Headless 360 — Salesforce's "Headless 360" transforms its platform for AI agent accessibility, moving beyond traditional GUI-centric SaaS.
• 🎧 Opus 4.5 Changed How Andrew Wilkinson Works and Lives — AI, particularly Opus 4.5 and Claude Code, empowers rapid, customized software development for both personal and professional applications.
• Opus 4.6 and ChatGPT 5.3-Codex Are Here and the Labs Are at War — Leading AI labs are converging on coding agents as the foundation for general-purpose knowledge work agents.
• The Debate Over Anthropic’s New Product: Price or Existential Dread? — AI agent pricing and capabilities are sparking an existential re-evaluation of traditional software development workflows and costs.
• 🗞️ Anthropic releases Claude Opus 4.8 on the same day as its $965B valuation round. — The AI frontier is rapidly expanding with new model capabilities, inference optimizations, and valuation milestones.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Matthew Berman.