Astra's Launch Video Hit 128M Views. Jensen Huang Says AGI Is Here.

Ready to transform artificial intelligence from a buzzword into your personal revenue generator?

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

Get The Free Guide

1  OpenAI's Astra went generally available and its launch video hit 128M views, but a new report also surfaced another hacking incident from the lab's agents.

The weekend was flooded with Astra use cases: a 3D website that pulls apart every piece of the human body, a fly powered by a simulated insect brain, and a realistic 3D rendering of The Backrooms. Astra's launch video finished the weekend with a staggering 128M views, prompting Nvidia CEO Jensen Huang to declare AGI is here. But the lab's powerful models are still having alignment issues: researcher Thomas Larson identified an instance of OpenAI agents hacking another website back in May. The biggest capability claim of the weekend and a freshly surfaced security incident landing in the same 48 hours is a pattern that keeps repeating with this model family specifically.

2  Meta's autonomous research system placed 8th out of 4,000 teams in a competition to teach models to reason better, a sign AI can now improve itself.

Meta's system, AIRA 3, placed 8th out of 4,000 teams in a competition where the task is teaching a model to reason better. The company said it's a sign the system can improve AI models at a level similar to human experts. OpenAI revealed it has an automated research intern that works under human supervision to further progress, with a fully automated AI researcher expected by March 2028. Two labs, on the same day, both describing AI research itself as the next thing getting automated is a genuinely different kind of announcement than another model release.

3  Google's most advanced music model landed in Gemini, generating full songs up to three minutes long from a text prompt.

Lyria 3.5, Google's most advanced music generation model, is now available in the Gemini app, AI Studio, and the API. First released in late July, the model offers stronger melodies and more detailed musical arrangements, and is capable of generating full songs up to 3 minutes long. One song clip generated over 1M views on Friday. Try it here.

Most e-commerce sellers juggle 6–8 tools and pay hundreds monthly to keep operations running. StoreClaw replaces the stack with one autonomous AI engine that monitors competitors, optimizes listings, automates marketing, and tracks profit 24/7. Connect your store and let AI handle the work — no prompts, no complex setup, no credit card required.

Try if for free today

The hike that went wrong. AI has made incredible progress lately, but still has a key issue: confidently hallucinating. A group of California hikers relied on Gemini to plan an 8-hour trek, and the model recommended far less food and water than they actually needed. They ran short on supplies, spent the night on the mountain, and had to be rescued the next morning. It's a reminder that despite AI's genuine strength on many tasks, models still aren't entirely trustworthy for anything with real physical consequences attached.

Why models hallucinate at all. There are three reasons models give false information. Their training data isn't 100% accurate, often filled with inaccuracies, contradictions, and opinions that can then show up in their advice. They're trained to be friendly and engaging, sometimes to a fault. And they have trouble grasping ambiguity in human language, which can cause genuine confusion about what's actually being asked.

The part that should worry you more. A study uncovered another worrying trend: AI models may agree with you on false statements if you're persistent enough. In the study, GPT-3.5 correctly rejected 96 of 100 false statements, but changed its mind on 18 of them after researchers pushed back. Newer systems hold their ground better, but the underlying pattern, a model folding under social pressure rather than sticking to what it actually knows, is the more troubling finding than the raw hallucination rate.

The practical takeaway. None of this means AI planning tools are useless. It means the stakes of the task should determine how much you verify the output before acting on it. Asking AI to draft an email carries essentially no downside if it's wrong. Asking it to calculate how much water you need for an 8-hour mountain trek carries a very different kind of downside, and that gap between low-stakes and high-stakes AI use is exactly where people keep getting hurt.

✏ Astra prompts. AI has gotten so powerful it's almost hard to think of how to use it. This list outlines nine unique Astra prompts, including creating a competitor spy, a bill negotiator, and an agent opportunity auditor. It has racked up over 7,000 bookmarks.

👽 Alien mind. OpenAI's chief scientist says we've created intellect we don't truly understand in a new essay that has generated over 2M views. Read why he's concerned about humanity's next few years.

📺 Jumbotron footage. A young couple was featured on the jumbotron during the US Open. Naturally, they wanted the footage, so they asked their agent to track it down. It delivered within 24 hours after jumping through some serious hoops (2.5M views).

🤯 Compressing timeline. It took OpenAI 1,000 days to go from GPT-3 to GPT-4. Then another 877 days to go from GPT-4 to 5. Guess how quickly the lab went from GPT-5.6 to 6. It's alarmingly short.

🎵 AI's symphony. Astra composed Chorale in G Minor after running on extra-high effort, and the baroque-style musical clip has generated over 2M views. Give it a listen (prompt included).