Coinbase is laying off nearly 700 employees - roughly 14% of its workforce - in what CEO Brian Armstrong calls an "AI-native" restructuring.
Every tech layoff now comes with AI branding. But Coinbase is being unusually explicit about it: artificial intelligence is replacing humans, not augmenting them.
According to Armstrong, engineers have demonstrated they can "use AI to ship in days what used to take a team weeks." Non-technical employees are now "shipping production code," and the company is moving toward organizing around "AI-native talent who can manage fleets of agents to drive outsized impact."
Translation: AI tools are good enough now that Coinbase thinks it can operate with significantly fewer people.
Is this real efficiency, or using AI as cover for cost-cutting during a crypto market downturn? Probably both. Armstrong admitted that cryptocurrency market conditions contributed to the decision, noting the company needed to become "lean, fast and AI-native."
The technology is genuinely capable. AI coding assistants can dramatically accelerate certain types of development work. Customer service chatbots can handle many routine queries. But there's a gap between "AI can help with this" and "we can replace 700 people with AI."
What happens when the AI can't handle edge cases? What happens when a customer has a complex issue that requires human judgment? What happens when the market conditions change and Coinbase needs to scale back up?
One Reddit user noted: "They're not replacing 700 workers with AI. They're expecting the remaining workers to use AI to do the work of 700 people. That's different."
That's the key insight. This isn't about AI autonomously performing these jobs. It's about Coinbase betting that with AI tools, fewer humans can do more work. The people who remain are expected to manage "fleets of agents" - meaning they're becoming AI supervisors rather than doing the work directly.
