OpenAI Assembling 'Superalignment' Squad 😳

Good morning, AI explorers!

Welcome to another edition of AI Authority, delivering all of the top groundbreaking AI news shaping our world into a bite-sized newsletter every weekday morning.

As always, we’ve got one daily AI headline, two exciting tools, and three pieces of content we love from the community for you today.

In this edition:

✍️ News: OpenAI’s Superalignment Squad Takes Shape
🛠️ Tools: Decktopus, TLDV
🎨 Creator Corner: A Legendary Musician’s AI Takes, Midjourney’s Jack Sparrow Improvement, Math Prompts

Here’s everything you need to know:

One top headline in the AI space to know:

🤖 OpenAI Forming Team to Prepare for Superintelligent AI

How far away do you think we are from a ‘superintelligent’ artificial intelligence that surpasses the capabilities of humans?

While superintelligence might sound like technology from a distant future (or completely science fiction), one major company is already preparing — and believes this reality is coming in the next decade.

OpenAI announced the formation of a new specialized team on Wednesday, led by Chief Scientist and Co-Founder Ilya Sutskever, to address the critical question:

How do we guide and control AI systems when their intelligence surpasses our own?

Right now, strategies for aligning AI lean heavily on human supervision — a problem when dealing with an entity potentially smarter than the supervisor.

"Currently, we don't have a solution for steering or controlling a potentially superintelligent AI, and preventing it from going rogue," OpenAI wrote in a blog post Wednesday.

"Our current techniques for aligning AI, such as reinforcement learning from human feedback, rely on humans’ ability to supervise AI. But humans won’t be able to reliably supervise AI systems much smarter than us.

The "Superalignment" team will be dedicated to addressing the technical challenges of controlling superintelligent AI over the next four years — and OpenAI is even dedicating 20% of its compute to the initiative.

The ambitious plan involves creating a "human-level automated alignment researcher," a system trained using human feedback that can aid in the evaluation of other AI systems.

Ultimately, the goal is to build an AI capable of performing alignment research — ensuring systems achieve the desired outcomes of humans without losing control.

And the stakes, according to OpenAI, are high.

“Superintelligence will be the most impactful technology humanity has ever invented, and could help us solve many of the world’s most important problems. But the vast power of superintelligence could also be very dangerous, and could lead to the disempowerment of humanity or even human extinction.”

While we’ll be rooting for humans to keep technology from destroying the world — we do have one suggestion for the superteam:

Two AI-powered platforms to add to your toolbox:

• Decktopus— AI-powered presentation generator (link)

• Tldv — Record, transcribe and summarize meetings and calls (link)

Three of our favorite tweets, threads, AI generations and more:

A great interview with musician will.i.am, who clearly understands how big of an impact AI is about to have on music:

A good look at how much Midjourney has improved in each version, through the lens of Captain Jack Sparrow! Via @linusekenstam:

An awesome prompt combination of math and M.C. Escher from @tivadardanka:

See you tomorrow!

If you haven’t already, subscribe to receive AI Authority in your inbox every weekday. And share it with an AI-curious friend!