Latest Update

When Is AGI Coming? Google's CEO Has an Answer

2026.05.21

Hi, it's the Carat team. 🤗
Google DeepMind's CEO publicly committed to an AGI timeline. OpenAI solved an 80-year-old math problem with AI, and Cursor launched an automation feature.

OpenAI Just Solved an 80-Year-Old Math Problem

OpenAI used AI to prove the 'plane unit distance problem' proposed by mathematician Erdős in 1946. For 80 years, the square lattice was believed to be optimal, but AI found a better arrangement and overturned that assumption.

The key point is that a general-purpose reasoning model did this, not a specialized math system. AI proposed a new approach to something math experts couldn't crack for decades.

AI isn't just finding answers faster. It's suggesting directions humans hadn't considered. Its role is shifting from 'tool' to 'co-researcher.'

📌 Today's Top 3

1️⃣ Cursor Now Automates Code Reviews

Cursor launched an automation feature. You can now automate repetitive tasks inside the agent workspace. Set up code reviews, test runs, and deployment prep in advance, and they run on their own.

Cursor is rapidly evolving from a coding tool into a full development platform. The automation feature cuts repetitive work and lets you focus purely on coding.

2️⃣ Google DeepMind CEO: 'AGI Starts Within 2-3 Years'

Google DeepMind CEO Demis Hassabis said in a post-Google I/O 2026 interview that AGI will begin in 2026-2027 and ramp up by 2029-2030.

He described AI's impact as '10x the Industrial Revolution at 10x the speed, so 100x total.' However, he also noted AGI won't be a sudden singularity but will unfold gradually. It's significant that Google's AI chief publicly named specific timelines.

3️⃣ Anthropic Asks: What Does 'Good Character' Mean for AI?

Anthropic released an 'AI Ethics Conversation Series.' It explores how to instill good character in AI models, featuring discussions with philosophers, theologians, ethicists, and constitutional scholars. The first topic: 'How is good character formed?'

Anthropic argues that making AI safer requires more than technical approaches. Humanity's accumulated ethical wisdom needs to be woven into AI development. Amid the performance race, asking 'what values does this AI hold?' first is a rare move.

🧪 Today's Prompt Tip

Floating Dog in the Sky Photo

Create a wide-angle photo of a dog from below, as if it's floating in the sky. This prompt went viral on Instagram @artemis.dal, racking up over 3,000 likes.

Here's a result we made on Carat

Generated with ChatGPT Images 2.0 on Carat

Paste the prompt below into Carat.

Use ChatGPT Images 2.0 Medium model with the prompt below to generate an image.

Create a photo of a dog (or cat) looking up, shot from below with the sky as the background. Use a wide-angle lens so the nose looks exaggeratedly large. Natural light, outdoor, clear sky.

Try it yourself. 👉 Create on Carat

That's all for today. See you tomorrow with more AI news. ☺️

Was this helpful? Share with your friends

Get daily AI trend updates