- Horizon AI
- Posts
- Top AI CEOs Debate What Comes After AGI 👀
Top AI CEOs Debate What Comes After AGI 👀
Anthropic Claude Opus 4.6 gets ‘Fast Mode’
Gina Acosta
February 08, 2026
In partnership with
Welcome to another edition of Horizon AI,
Google’s Demis Hassabis and Anthropic’s Dario Amodei recently sat down to debate the rapid arrival of Artificial General Intelligence and how it will reshape our world.
Let’s jump right in!
Read Time: 4.5’ min
Here's what's new today in the Horizon AI
Chart of the week: Which Jobs Are Most Exposed to Generative AI?
Anthropic Launches Fast Mode for Claude Opus 4.6
AI Findings/Resources
AI tools to check out
Video of the week: Is AGI arriving in 2026? The Giants Speak: A Roadmap to the AGI Era
TOGETHER WITH OUTSKILL
Build Real AI Automations Next Weekend
Most AI advice stays theoretical. Outskill's 2 Day AI Mastermind is 16 hours of live, hands-on sessions where you build actual AI workflows, automations, and revenue-generating use cases.
Rated 4.9/5 on Trustpilot.
What you'll walk away with:
10+ AI tools and automations you can use immediately
End-to-end agentic workflows you build during the sessions
The 2026 AI Survival Handbook
$5,000+ in bonus resources, including a Prompt Bible and AI Monetization Roadmap
📅 Feb 14-15 | 🕐 10 AM-7 PM EST | Free
Chart of the week
Which Jobs Are Most Exposed to Generative AI?
Microsoft analyzed 200,000+ Copilot conversations to estimate how exposed different jobs are to generative AI based on task coverage, completion, and overall applicability.
Language, research, writing, and customer-facing roles rank highest, with interpreters and translators topping the list due to how often and how well AI handles their tasks.
High exposure signals strong AI assistance potential, not automatic job replacement, with physical and hands-on roles remaining far less affected.
AI News
ANTHROPIC
Anthropic Launches Fast Mode for Claude Opus 4.6
Anthropic has introduced a new Fast Mode for Claude Opus 4.6 that prioritizes speed over cost.
Details:
The company says responses are about 2.5× faster while maintaining the same output quality, though it costs up to 6× more than standard Opus 4.6 pricing, making it clear it's not meant for the average user.
According to Anthropic, the mode is built for urgent, high-stakes projects where you need speed for interactive work like rapid iteration or live debugging, and it can be toggled off when cost matters more than latency.
Fast Mode can be toggled on in Claude Code with /fast and works across Cursor, GitHub Copilot, Figma, and Windsurf.
It's available now for Claude Code users with extra usage enabled, with plans to expand API access down the line. Interested developers can sign up for a waiting list.
AI Findings/Resources
🔮 Marc Andreessen foresees a future where generalists dominate the AI era
📈 Grok just passed DeepSeek for the first time, making it the third most visited Gen AI tool
AI Tools to check out
👤 ConnectMachine: AI Agent for digital business cards & contact management for the silent elites who prefer private, purposeful networking and meaningful connections.
🛠 Xcode: Develop, test, and distribute apps for all Apple platforms.
🤝 Crisp: AI customer support platform.
🦾 Relay.app: The easiest way to create a team of AI agents that work for you across Gmail, Notion, HubSpot and hundreds of other apps.
🛒 Genstore.ai: Describe your idea in a single prompt and let AI agents spin up a live, sell-ready store in minutes.
Video of the week
Is AGI arriving in 2026? The Giants Speak: A Roadmap to the AGI Era
The leaders of the world’s two most prominent AI labs just shared a stage, and their predictions for the next two years are a wake-up call for every industry. While they differ slightly on the "how," both agree that we are no longer talking about decades, we are talking about months.
The Shrinking Timeline ⏳
Dario Amodei (Anthropic) doubled down on his bold prediction: we could see AI models capable of doing everything a human can do by 2026 or 2027. He believes we are roughly 6 to 12 months away from a "closed-loop" system where AI begins to write its own code and conduct its own research. Demis Hassabis (Google) is slightly more cautious, sticking to a 50% chance of reaching this milestone by the end of the decade, noting that while AI excels at math and coding, "creative" science, like asking the right questions, remains a harder hurdle.
The Jobs Crisis and "Capability Overhang" 💼
The conversation took a sobering turn regarding the labor market. Amodei warned that up to half of entry-level white-collar jobs could be displaced within one to five years. Hassabis noted that while new, more meaningful roles will eventually emerge, there is currently a "capability overhang", meaning today’s AI tools are already more powerful than most people know how to use. His advice to the next generation? Become "unbelievably proficient" with these tools now to leapfrog traditional career paths.
Managing the Risks 🛡️
Perhaps the most striking moment was Amodei’s "Battle Plan" for surviving our "technological adolescence". He argued that as we build "machines out of sand," the risk of misuse by bad actors or nation-states is a crisis that requires all of our focus. Both leaders expressed concern that society hasn't built the institutions needed to handle the massive wealth and productivity shifts AGI will bring.
That’s a wrap!
We'd love to hear your thoughts on today's email!Your feedback helps us improve our content |
Not subscribed yet? Sign up here and send it to a colleague or friend!
See you in our next edition!
Gina 👩🏻💻