AI Safety

AI Experts Now Think Superintelligence Could Arrive by 2028

AI Experts Now Think Superintelligence Could Arrive by 2028

⚡Top researchers warn we may have only 1-2 years to prepare for AI smarter than humans.

Deep Dive

A recent report from METR (a research group that studies AI progress) found that AI systems are getting better at doing AI research itself, which could speed up AI development dramatically. Some forecasters now believe we could reach superintelligence—AI that is much smarter than humans—in as little as 1-2 years. This is a big deal because superintelligent AI could solve major problems like disease and climate change, but if it's not properly controlled, it could also pose serious risks.

The report has caused a stir among AI experts. Ryan Greenblatt, a researcher, extrapolated the findings and suggested that superintelligence might arrive around July 2027, which is faster than the already aggressive 'AI 2027' scenario. The AI Futures Project released an emergency update, and now some experts put the median date for superintelligence around March 2028—about a year and a half away. If superintelligence arrives within 6-12 months, one estimate puts the chance of an 'AI takeover' (where AI systems end up beyond human control) at roughly 75%. That means unless we slow things down, we face a 1-in-4 to 1-in-3 chance of a dangerous loss of control within a year.

To prevent this, experts are calling for 'pacing' or 'pausing' AI development—essentially slowing down the race to superintelligence. They suggest three main paths: lab advocacy (convincing AI companies to slow down), lobbying governments (especially California and the US federal government) to impose regulations, and technical work to prepare for a global pause, like third-party assessments to build trust between countries. These efforts are currently underfunded and neglected, according to the author. The goal is to buy at least a year of extra time to ensure AI is developed safely.

For the general public, this means the AI revolution might be coming faster than expected. While superintelligence could bring amazing benefits, the risks are real. Experts are urging immediate action to ensure we don't lose control. The good news is that there are concrete steps we can take, but they require rapid and coordinated effort from researchers, companies, and governments.

Key Points
  • AI progress is accelerating: a new report suggests superintelligence could arrive by 2028, much sooner than previously thought.
  • Without safety measures, there's a 1-in-4 to 1-in-3 chance of a dangerous AI loss of control within 12 months of superintelligence.
  • Experts are calling for immediate action to slow down AI development and prepare for a global pause.

Why It Matters

Faster AI progress means we need to act quickly to ensure it's safe and beneficial for everyone.

🔒 FREE VIP ACCESS

Read the full breakdown & get daily AI executive intelligence

Join like-minded builders, researchers, and AI professionals who read The Neural Feed daily. 100% free. No spam.

🔔 BREAKING RELEASE RADAR INSTANT ALERTS • 1-CLICK PAUSE

Get instant breaking AI release alerts by topic

Choose the AI labs, models, or chip tracks you follow. Receive fast alerts only when something major ships — plus our free daily morning briefing.

Includes full article unlock • Only major releases • Unsubscribe anytime
🎁 FREE EXECUTIVE DOWNLOAD ISSUE #21

Download This Week's 8-Page Executive AI Briefing (PDF)

Our curated executive intelligence packet: in-depth breakdowns of frontier models, market movements, plus full-color AI comics.

  • ✓ 8-page formatted executive PDF briefing
  • ✓ Frontier AI models & technical breakthroughs
  • ✓ Full-color satirical AI comic strips
Instant 1-click download • 100% free • Unsubscribe anytime
📬 Get the top 10 AI stories daily