← Back to archive
Spotify's best devs ditched coding for AI

Image via Unknown

Friday, February 13, 2026

Spotify's best devs ditched coding for AI

OpenAI's lightweight Codex is racing ahead with Cerebras (yikes, the speed wars are real), while Anthropic just raised $30B at a $380B valuation on the back of $14B in enterprise revenue, proving AI cash flows are no longer theoretical. Meanwhile, Google's upgrading Gemini with Deep Think for serious reasoning chops, but here's the kicker: Spotify's best engineers haven't coded since December, and OpenEnv just showed us that real-world agents are still fumbling the multi-step plays that matter. So here's the question: if AI agents keep failing in production, who's actually buying this stuff?

Top Stories

1
This is for the speed addicts

OpenAI and Cerebras launched Codex-Spark, a low-latency coding tool powered by custom hardware, demonstrating the industry's shift toward specialized chips for faster AI inference in real-time applications.

openaicodexinferencehardware
2
Spotify's Best Developers Haven't Written A Line Of Code Since December

Spotify's top developers have stopped writing code directly, instead using AI to automate development via an internal 'Honk' system, enabling remote real-time deployment and unprecedented feature velocity while building defensible AI datasets around music preferences.

ai-codingllmclaudeenterprise-ai-adoption
3
Real-World Agent Evaluation

Hugging Face

OpenEnv standardizes real-world AI agent evaluation beyond simulations, revealing that agents struggle with multi-step reasoning, ambiguity resolution, and execution reliability despite succeeding at individual tool calls—critical insights for building production-ready agent systems.

agentstool-useevaluationopen-source
4
Anthropic Raised $30B at $380B Valuation

Anthropic

Anthropic's $30B Series G fundraise at $380B valuation reflects explosive enterprise adoption of Claude, with $14B run-rate revenue and Claude Code accounting for $2.5B, establishing the company as the clear leader in enterprise AI and agentic coding.

anthropicllmfundingenterprise-ai
5
Gemini 3 Deep Think Upgrade

Google

Google upgraded Gemini 3 Deep Think to achieve state-of-the-art reasoning across math, physics, and engineering domains, while expanding access through API availability for researchers and enterprises seeking to solve complex, unstructured problems.

geminigooglellmreasoning

Keep Reading

Industry Voices

Enjoyed this issue?

Get daily AI intel delivered to your inbox. No fluff, just the stories that matter.