Sent to 16 readers
Race to AGI Daily Digest - Wednesday, January 21, 2026
Share:
TLDR
Humans& raises a massive $480M seed to build a long‑horizon AI lab, deepening the AGI lab race.
Emergent secures $70M to scale its ‘vibe‑coding’ platform, pushing agentic coding tools toward the enterprise frontline.
AI security stack forms as Dam Secure funds code‑hardening tools and HackerOne launches an AI Safe Harbor program.
Public AI and chip leaders slide again even as private capital doubles down on AGI labs and agent platforms.
The Full Story
Following Monday’s Claude Cowork launch and Tuesday’s AI sell‑off, you’d expect money to cool a bit. Instead, today looks like the opposite: public AI names are bleeding red, while private AGI bets are catching fire.
On screens, the old guard is hurting. Oracle, Broadcom, IBM, TSMC, and NVIDIA are all down around 4–6%, another day where the “AI trade” feels crowded and jittery. But behind the scenes, the AGI lab race is accelerating.
Humans& just raised a huge $480M seed to build a long‑horizon AI lab. Long‑horizon here means models that can plan and reason over months or years, not just respond to one prompt. That slots directly into the "AGI Capital Concentrates Around a Few Labs" thread we’ve been following, especially with backers including NVIDIA and Google. You can dig into what Humans& is actually building here ->.
Building on Monday’s “AI coworker” theme, Emergent is going after how we write software at all. Its ‘vibe‑coding’ platform is about telling an AI the feel and intent of what you want, and letting agents handle the boilerplate. Emergent just locked in a $70M Series B from names like Khosla Ventures and SoftBank, on top of broader coverage of its India‑focused push. Catch the Series B details here -> and the regional funding lens here ->.
Now, if agents are hitting the enterprise frontline, who secures the mess they create? Dam Secure raised $4M to harden AI‑generated code before it ships to production here ->. And HackerOne’s new Good Faith AI Safe Harbor, launched with Anthropic in the mix, gives security researchers clearer protection when probing AI systems for bugs here ->.
Meanwhile, the background hum is all about specialized hardware. The ongoing NVIDIA–Groq talent and technology deal narrative keeps pointing toward inference‑heavy, agent‑style workloads as the next battleground track that story ->, and NVIDIA’s central role in nearly every big lab round is hard to miss company profile ->.
So by midweek, the picture is clearer: agents are stepping into real work, capital is clustering around a few ambitious labs and tools, and a new AI security stack is racing to keep up.
Get This Delivered Daily
Join thousands of AI professionals who start their day with Race to AGI.