← All briefings
📰

subtl daily briefing

Share𝕏in
AllAICryptoSaaSFinanceStartups

Good morning, founders and operators. OpenAI dropped GPT-5.5 with full agentic capabilities and a $5/M token price tag, the same week Anthropic quietly crossed a $1 trillion valuation on secondary markets — making this the most consequential week for AI market structure in months. Meanwhile, a $293M DeFi hack rattled institutional confidence in open protocols, and Meta is cutting 8,000 jobs to fund its AI infrastructure bet.

In today's briefing

  • 1.GPT-5.5 launches, Anthropic hits $1T
  • 2.Meta cuts 8,000 jobs for AI bet
  • 3.Kelp DAO $293M hack shakes DeFi
  • 4.Bitcoin surges 24% amid geopolitical chaos
  • 5.The 'Gen Marketer' AI hiring archetype arrives
  • ⚡Quick hits on other news
Latest Developments
AI

🤖GPT-5.5 Goes Live and Anthropic Quietly Crosses $1 Trillion

The Rundown: OpenAI launched GPT-5.5 with advanced agentic capabilities and a 1M token context window, while Anthropic surpassed OpenAI's own valuation on secondary markets at $1 trillion.

The details:

  • ●GPT-5.5 scores 82.7% on Terminal-Bench 2.0, is priced at $5/M input tokens, and supports autonomous operation across email, spreadsheets, calendars, and Slack via OpenAI Workspace Agents (free until May 6 for Business/Enterprise plans)
  • ●Anthropic crossed a $1 trillion valuation on Forge Global, driven by surging Claude Code adoption and limited share availability — surpassing OpenAI on secondary markets
  • ●DeepSeek V4-Pro launched simultaneously with 1.6 trillion parameters and a 1M token context window, confirmed compatible with both Nvidia and Huawei chips amid US export restrictions
  • ●Alibaba's Qwen3.6-27B outperforms its own 397B model on coding benchmarks (SWE-bench 77.2 vs 76.2), runs on just 18GB VRAM, and is Apache 2.0 licensed
Why it matters: This week crystallized a new competitive dynamic: the frontier AI race is no longer just about model quality but about agentic utility and pricing. GPT-5.5's $5/M token price and Workspace Agents integration signals OpenAI is targeting enterprise workflow automation directly. For founders, the more alarming signal is from TLDR Founders — a Claude integration replicated 95% of a vendor's AI features at 15% of the token cost, causing a customer to slash their SaaS renewal by 45%. If you're selling software with any AI feature layer, your renewal conversations are about to get harder.

📰 Source: TLDR, AlphaSignal, Techpresso

Share𝕏in
AI

🔪Meta Axes 8,000 Jobs to Fund $135B AI Infrastructure Blitz

The Rundown: Meta is cutting 8,000 employees and cancelling 6,000 open roles to redirect capital toward up to $135 billion in AI infrastructure spending this year.

The details:

  • ●Meta layoffs affect ~10% of staff (~8,000 jobs) effective May 20, with 6,000 open roles also cancelled — all savings redirected to AI compute
  • ●Google reports 75% of its code is now AI-written, with API token usage growing 60% in Q1 driven by agentic AI adoption
  • ●S&P 500 companies have reported only $300M in collective AI productivity gains, yet Anthropic earns roughly that amount in token sales every 1.5 days — a stark gap between enterprise adoption and frontier AI revenue
  • ●The 'tokenmaxxing' trend has employees at Meta, Microsoft, and Salesforce wastefully burning AI tokens for internal leaderboard rankings, causing financial waste and system outages
Why it matters: Meta is making the most explicit corporate bet yet that AI tokens are more cost-effective than human headcount — and the market is rewarding it with near-all-time-high stock prices. For founders and operators, this is both a playbook and a warning: companies that can credibly show AI is replacing labor costs will get re-rated by investors, while those that can't articulate the ROI will face pressure. The tokenmaxxing dysfunction is a real operational risk worth watching — unchecked AI compute costs can quietly become material.

📰 Source: Techpresso, MIT Technology Review, The Breakdown

Share𝕏in
⚡

Everything else in the news today

Cognition AI (maker of Devin) is in talks to raise at a $25B valuation — TLDR→
DeepSeek is in talks to raise its first funding round at a $20B valuation backed by Tencent and Alibaba — TLDR→
Z.ai released GLM-5.1, a 754B parameter open-weights MoE model that runs agentic coding loops for up to 8 hours, topping the Artificial Analysis Intelligence Index among open models at $1.40/$4.40 per million tokens — The Batch→
Band emerged from stealth with $17M to build a universal orchestration layer enabling AI agents across different frameworks and clouds to discover and collaborate in real time — TLDR→
The Bitwarden CLI npm package was briefly compromised via a malicious credential stealer exploiting a Checkmarx GitHub Action — rotate your npm tokens, SSH keys, and cloud credentials now — TLDR InfoSec→
The Zealot AI multi-agent system autonomously breached a GCP environment — from network recon to SSRF exploitation to BigQuery data exfiltration — without human intervention — TLDR InfoSec→
Microsoft issued an emergency patch for CVE-2026-40372 in ASP.NET Core on macOS/Linux — admins must upgrade, rotate DataProtection key rings, and revoke all long-lived tokens — TLDR InfoSec→
A worm-like npm supply-chain attack targeting Namastex Labs packages is stealing credentials and self-propagating to PyPI — rotate secrets immediately — TLDR DevOps→
Project Eleven awarded quantum cryptographer Giancarlo Lelli 1 BTC for cracking a 15-bit elliptic curve key, raising fresh alarms about Bitcoin's post-quantum security — Bankless→
ChatGPT Images 2.0 is now live across all major Figma products including Design, FigJam, Slides, and Weave — TLDR Design→
Instagram is testing a standalone app called Instants for unedited, once-viewable disappearing photos to compete with Snapchat and BeReal — TLDR Design→
Apple is rumored to launch a foldable iPhone Ultra capable of running two apps side-by-side alongside the iPhone 18 lineup — TLDR Design→
Kubernetes v1.36 shipped with 70 enhancements including fine-grained kubelet API authorization and user namespaces, while deprecating the vulnerable externalIPs field — TLDR DevOps→
Pyroscope 2.0 reduces symbol storage by 95% and eliminates write-path replication, making continuous profiling dramatically cheaper at scale — TLDR DevOps→
AI systems now apply a 'bland tax' that filters generic content from AI search results — brands without a distinct voice are losing attribution entirely — TLDR Marketing→
📰Today⚡Feed📡Signals💰Capital