🫴 OpenAI bribing publishers?

The Hompage of AI

Welcome to AI Tool Report!

Friday’s top story: A leaked slide deck has revealed how OpenAI is wooing publishers into letting them use their content to train its AI models. 😳😳

🌤️ This Morning on AI Tool Report

1. 🫴 Is OpenAI bribing publishers?

2. 🔒 Reddit locks down its content

3. ⚡OpenAI steals Google’s thunder (again)

4. 🏷️ TikTok auto-labeling AI content

Read Time: 5 minutes

STOCK MARKETS

PUBLISHING

🫴 Is OpenAI bribing publishers?

Our Report: A leaked slide deck has revealed how OpenAI is wooing global publishers—like The FT and Le Monde—to partner with them so they can use their content to train their GPT models and legally display information, without getting sued.

🔑 Key Points:

• The deck revealed that OpenAI has an initiative—the “Preferred Publishers Program” (PPP)—that negotiates deals on a per-publisher basis, rather than structuring its deals uniformly.

• According to the deck, the PPP is just available to “high-quality editorial partners” and promises them priority placement, “richer brand expression”, and more prominent links.

• PPP members get a standard rate for allowing OpenAI to use its data and a variable one which is dependent on how manyChatGPT users engage with their displayed or linked content.

🤔 Why you should care: This revelation has sparked concern that commercial models, like ChatGPT, will now start displaying biased content, based on ad spend, a little like Google, perhaps.

Together with Hedonova

Hedonova has Outperformed the S&P 500 by Over 200% 📈

Investing in alternative assets can often be complex and illiquid, making them challenging to fund. That's where Hedonova comes in, making it easy to invest in alternative assets.

Since late 2019, Hedonova has achieved a remarkable feat, outperforming the S&P 500 by more than 200%.

Hedonova provides a simple yet powerful proposition: It serves as a single entry point to a diverse portfolio of alternative assets, including media royalties, pre-IPO startups, wine, and fine art.

This SEC-regulated fund has captured the spotlight, securing a multi-strategy hedge fund award at the 2023 Hedgeweek European Awards.

With a low minimum of $10k, easily start investing in alternative assets at Hedonova

1. Vizly analyzes Excel sheets, CSVs, JSON files, PDFs, and images

2. HitWit collects and analyzes ideas from high-quality sources

3. Retinello creates customizable educational experiences

4. illumidesk generates e-learning lessons with AI

5. Coursnap is for learning YouTube courses

LICENSING

🔒 Reddit locks down its content 

Our Report: Reddit has announced a public content policy designed to prevent commercial companies from using its user-posted content and data, without an agreement.

🔑 Key Points:

• The policy states that those in non-commercial or research outlets can still use its data, but those who plan to use it for commercial gain—such as AI training—will have to pay. 

• It establishes that partners can’t use Reddit’s content for ad targeting or to conduct “background checks, facial recognition, government surveillance, or help law enforcement.”

• It also clarifies that Reddit will not sell private or non-public information—such as messages, emails, or mobile numbers—to data licensees, which include AI companies.

🤔 Why you should care: Reddit has published this policy because it’s trying to balance its business need to license its content to big tech companies (it already has a licensing agreement with Google worth $60M) with protecting its user's privacy rights.

⌛ JUST 24 hours left to get AI For Professionals for $379 (31% saving) ⌛

SEARCH ENGINES

⚡OpenAI steals Google’s thunder (again)

Our Report: OpenAI’s rumored ChatGPT search engine is expected to be announced on Monday, May 13th, just one day before the king of search engines, Google, starts its annual developer conference (Google I/O)—where it's expected to make some big AI announcements—completely stealing their limelight.

🔑 Key Points:

• The ChatGPT-powered search engine was originally expected to be released on May 9th, but reports say OpenAI changed its plans, to gazump Google’s important announcements.

• It’s thought it will directly challenge Google’s Search Generative Experience by collating information from different sources and giving users summarised answers to their queries.

• This isn’t the first time OpenAI has stolen Google’s limelight: When Google announced Gemini was replacing Bard, a week later, OpenAI announced their text-to-video generator Sora.

🤔 Why you should care: The battle for limelight aside, studies show that while people are curious about AI-powered search experiences, they don’t always trust the information they’re presented with, meaning OpenAI has its work cut out if it wants to take the king of search down.

ETHICS

🏷️ TikTok auto-labeling AI content

Our Report: While TikTok already labels content made with TikTok AI, it’s started automatically labeling AI-created content made on other platforms–like OpenAI’s DALL-E–with “AI-generated” to notify users that it was made with AI.

🔑 Key Points:

• They’re doing this by implementing Content Credentials (made by the Coalition for Content Provenance and Authenticity (C2PA), which Microsoft, Adobe, and OpenAI are already doing.

• Content Credentials attach data (like where and how the AI-generated content was made or edited) to content that TikTok can then use to recognize and label AI-generated content.

• TikTok already asks creators to disclose whether content has been made with AI, but this takes the pressure away from the creator and is another way to ensure AI content is labeled.

🤔 Why you should care: TikTok has also announced that, in the coming months, they will also start attaching Content Credentials to content created using TikTok AI, showing a commitment to ensuring that AI content made in TikTok is also labeled when posted on other platforms.

🍔 Quick hits

• Mistral AI, is raising funds at a $6B valuation, with DST, General Catalyst, and Lightspeed Venture Partners, interested.

• Anthropic has defended its decision to partner with both Google and Amazon, claiming it gives them more independence than if they just had one.

• OpenAI has hinted that ChatGPT-5 will have more personality and be more of a buddy than a prompt engineer.

• So far, the NY Times has spent over $1M in its copyright infringement lawsuit against OpenAI but insists that this hasn’t affected its operations.

We read your emails, comments, and poll replies daily.

Hit reply and let us know what you want more of!

Until next time, Martin & Liam.

P.S. Don’t forget, you can unsubscribe if you don’t want us to land in your inbox anymore.