🤖 ChatGPT Policing Itself

What's up? You're reading Inclined AI. Unlike “Game of Thrones,” we hope to stay entertaining and never phone it in. We can promise you that.

Here's what we’re watching from the wall:

• ChatGPT dreams of moderation

• Google Chrome gets an update

• The NYT is planning on suing OpenAI

• Eric Schmidt is planning an AI moonshot

Be a sponsor of our newsletter.

In an odd twist, OpenAI found the perfect content moderator for their tooling: ChatGPT.

Yes, the LLM behind the chat interface is moderating itself and doing a solid job. The Head of Safety at OpenAI, Lilian Weng, even believes ChatGPT can monitor content for other websites. 

It’s a massive step forward for AI technology.

But the setup is far from perfect. Experience content moderators do a better job sorting and handling issues that fall in the gray area—as with any industry, seasoning matters.

Still, ChatGPT showcases better abilities than a recruit, and that’s a promising development for AI. Why? Picture a future where AI can help comb through training data and moderate new models.

That spares humanity from coming face to face with the worst things the internet has to offer.

That issue is a problem that companies like OpenAI need to address today, and this solution can make waves now. No one wants them to pay cents on the dollar to people with no other option but to look at hateful, terrible content all day.

Of course, we can’t help but wonder how this affects content in the future. Trusting a clustered black box with what we watch and read is a leap of faith.

This truth begs the question: why did we even trust humans in the first place?

Humans are biased and flawed. They only run content moderation teams because AI wasn’t up for the task. Perhaps ChatGPT is the mana from heaven that gives tech companies an out.

You can’t argue with an AI trained on a clear set of policies, whereas you can accuse humans (or humans that make algorithms) of acting out of turn.

This setup would create less gray space and is worth the trade-off of letting AI prune our content intake.

Life-Changing Concepts → gives you an unfair advantage in business using mental models

Master ChatGPT → check out the ChatGPTricks e-book that takes you from zero to hero

Inclined AI Premium Newsletter → get a free 7-day trial on us when you subscribe today

The Strategy Deck → Find out what AI platforms will give you a competitive advantage and find relevant, actionable information in every edition.

The rumors were true. Google Chrome got an upgrade. 

The new version includes Search Generative Experience while Browsing. That long ass name is what Google calls it. I call it a much-needed feature and another step towards matching Bing’s offering.

In fact, I didn’t read the article about the new tool. I summarized it with SGE while Browsing, and it worked great.

You can tell Google thought this through based on the details.

They baked the primary search benefits into the interface, and the AI is not disruptive. Google frames it as an easy option to access, and it doesn’t clutter the page with repetitive information.

This new SGE update will change user behavior. Why would you copy and paste an article into ChatGPT when the summary works fine? Why would you download Microsoft’s browser now when Chrome offers the same features?

Bing Chat is on the ropes, and Bard is on the come-up.

It feels like OpenAI, and Microsoft woke a sleeping giant. Sure, said giant moved at a glacier pace, but it’s up and running. Gemini is coming soon, and their AI tooling continues to improve.

Try it out and let me know what you think. For my money, I think it’s the future, and I find myself using ChatGPT less and less now that browsing is gone.

🌟 The aggressive search for GPUs is ramping up globally

🌕 Eric Schmidt, Google’s ex-CEO, is launching an AI science moonshot

💼 The New York Times is considering piling onto the OpenAI lawsuit parade

🗞️ The Associated Press defines new general standards for sourcing AI information

🚫 An Iowa schoolboard used AI to help them decide what books to ban

📷 Google Photos adds a new generative AI feature that will name and organize collections for you

🎓 Academic journals don’t know how to detect or address the rise of AI usage in their publications

📢 This marketing toolkit will help you unleash the full potential of AI

• SayData - build self-serve customer-facing analytics (link)

• Human Generator - create people w/ AI in real-time (link)

• insightIQ - helps brands get maximum ROI (link)

• Wavel - high-quality voice clones w/ just 60 seconds of audio (link)

• Genki - automate your SEO growth (link)

• Inbox Chat - ChatGPT as your personal email assistant (link)

• Vanna - Python-based AI SQL agent (link)

• IdeaScore - idea testing & refinement (link)

Turtles That Go Fast & Furious (But Also Care About Family) The Movie

Someone get Warner Bros. on the phone and let them know I have a pitch!

_{- That’s it for today. I hope you enjoyed the latest edition of inclined.ai - Davis.}