← All briefings
📰

subtl daily briefing

Share𝕏inEmail
AllAICryptoSaaSFinanceStartups

Good morning, tech enthusiasts. This morning’s briefing brings a sobering look at AI safety, with researchers and lawmakers sounding alarms about existential risks. We also have breakthroughs from DeepSeek and OpenAI, a major crypto investment from Nasdaq, and savvy marketing moves from top brands.

In today's briefing

  • 1.AI safety warnings go mainstream
  • 2.DeepSeek and OpenAI push AI boundaries
  • 3.Nasdaq backs Kraken as stablecoins gain traction
  • 4.Brands rethink sponsorships and social series
  • ⚡Quick hits on other news
Latest Developments
AI

⚠️AI Safety Warnings Go Mainstream with Extinction Risk and Political Action

The Rundown: AI safety warnings are escalating as researchers resign and lawmakers move to regulate superintelligence.

The details:

  • ●Anthropic researcher Jacob Coxon resigned with a viral warning that AI companies are gambling with lives by racing to superintelligence.
  • ●Anthropic alignment lead Evan Hubinger stated there is a >10% chance AI kills all humans in the next decade and that the company has no alignment plan.
  • ●Sens. Sanders and Hawley are introducing bills and launching probes into catastrophic AI risks.
  • ●OpenAI also reportedly issued its own warnings about existential risks from AI.
Why it matters: For founders and investors, this signals a regulatory shift that could reshape AI development timelines and risk management. Companies building AI must now account for potential government scrutiny and align with emerging safety standards or face reputational and legal risks.

📰 Source: Casey Newton, TLDR, The Neuron

Share𝕏inEmail
AI

🤖DeepSeek and OpenAI Push AI Boundaries with New Model Milestones

The Rundown: DeepSeek's 552B model outperformed its Pro variant before being deprecated, while OpenAI claims progress on a second Millennium Prize problem.

The details:

  • ●DeepSeek's 552B model beat its own Pro model in benchmarks, then the company removed it, suggesting a rapid iteration cycle.
  • ●OpenAI reported substantial progress on a second Millennium Prize problem, rumored to be the Hodge Conjecture, following its 10,000-agent Navier-Stokes run.
  • ●These developments highlight accelerating capabilities in open-source and frontier AI models, with potential applications in math and science.
  • ●The rapid pace of innovation is leading to models that are both more powerful and shorter-lived, forcing businesses to adapt quickly.
Why it matters: AI capabilities are accelerating faster than expected, creating both opportunities and risks for businesses building on these models. Startups that can leverage the latest breakthroughs stand to gain a competitive edge, while those relying on older models may find themselves obsolete sooner.

📰 Source: AlphaSignal, The Neuron

Share𝕏inEmail
⚡

Everything else in the news today

MIT Technology Review named under-35s shaping the future of biotech, highlighting young innovators.→
Anthropic said it disrupted five biological-misuse cases that could support weapons development.→
UMG and ElevenLabs partnered on a licensed AI remix platform; Suno released v6 with 5x faster generation.→
Oracle cloud revenue surged 121% in the latest quarter, per TLDR.→
Microsoft is targeting Salesforce with new AI features to compete in the CRM space.→
Away's Actually Away series achieved a 78% engagement rate, outpacing typical social media benchmarks.→
Lob reports 9 in 10 customers act on personalized direct mail, proving the channel's effectiveness.→
📰Briefing📈Intelligence📚Library