🤓 This Week in AI
Plus ChatGPT’s Helpful Yet Hidden Search Feature
Happy Friday!
This week, learn how to search your ChatGPT conversation history, and discover a new app that can transform your daily commute or walk into productive reading time.
But first, here’s what you need to know about AI this week:
Video game performers are going on strike after talks with video game companies stalled over the use of AI.
Following plagiarism accusations, Perplexity, an AI search engine, launches a program to share ad revenue with publishers like Time, Fortune and Entrepreneur. Publishers featured in queries will receive a "double-digit percentage" revenue share, a free one-year Enterprise Pro subscription and access to developer tools.
Unlike OpenAI's ad-free SearchGPT and Google's ad model, which doesn't share revenue, Perplexity’s strategy aims to set up long-term relationships and infrastructure to pay select publishers for their content. “But it doesn’t answer a lot of questions, like what that means for publishers that aren’t getting a check.”
If you work in media or publishing, I highly recommend reading the full article to better understand the nuances.AI music startup Suno is challenging a lawsuit from major record labels, arguing it aims to stifle competition. Filed by the Recording Industry Association of America (RIAA) on behalf of Universal Music Group, Warner Music Group, and Sony Music Entertainment, the lawsuit alleges that Suno and competitor Udio trained their AI models on copyrighted recordings without permission.
Suno argues that their practices fall under fair use - a legal defense that allows for the use of copyrighted material in the creation of a “meaningfully new or transformative work”. They also emphasize that their AI tool, which generates music from simple text prompts, democratizes music creation, making it accessible to a wider audience.
The record companies argue that “the startups cannot claim fair use, saying that the outputs of Udio and Suno are meant to replace real recordings, that they are generated for a commercial purpose, that the copying was extensive rather than selective, and finally, that the resulting product poses a direct threat to labels’ business.”Without notifying users, Twitter/X has automatically activated a setting that allows the company to train its Grok AI on users’ posts. The good news is that you can switch it off and also delete your conversation history with the AI (which I recommend.) Here’s how.
Google DeepMind's new AI systems achieved a silver medal performance at the International Mathematical Olympiad. This is impressive because AI systems often struggle with complex math problems. Just 18 months ago, experts thought it would take 5-7 years for AI to reach this level. Many of these problems require creativity, reasoning, and problem-solving abilities once considered uniquely human.
A new study found that mentioning AI in product descriptions lowers emotional trust and reduces purchasing decisions, especially for “high risk” products. This suggests a growing backlash and disillusionment with the tech.
💡 Here’s my take : This is just one study…..But AI is an understandably an emotionally charged topic and presents complex consumer perception and ethical challenges. A deep understanding of psychology and consumer attitudes is key to anticipating and addressing reactions and developing effective brand and communication strategies.The New York Times used AI to better match advertisers with relevant stories—and it’s performing better than non-AI techniques. The gen-AI-powered ad tool, called BrandMatch, uses large language models to decode an advertiser’s brief, uncover more nuanced audience segments, and match that brief to the most relevant articles to run ads against, allowing for more granular targeting.
Meta scraps celebrity chatbots that fell flat with users, but will roll out AI studio, a tool that allows users to create and share personalized AI characters. Instagram creators will be able to use the chatbots as an “extension of themselves” to handle common DM questions and story replies. And they’re sharable across Meta’s ecosystem (Instagram, Facebook and WhatsApp.)
Meta says it built transparency into its AI, making it clear in chats that users are talking to an AI, not a human being. When users send a message to an influencer’s chatbot, Meta AI’s response will include a warning: “[Influencer Name] has AI auto replies turned on. Some messages generated by AI may be inaccurate or inappropriate.” AI Studio is free to use but currently only available in the US without any additional details on planned expansion.
I’m not bullish on influencer doppelgangers, as they strip the experience of authentic connection (or at least the perception of it) for users, but am eager to know how the masses react….Taco Bell is set to expand the use of AI voice technology in its drive-thrus to hundreds of US locations by the end of this year, even after McDonalds decided to stop its own AI drive-thru pilot
INSIDER TIPS
You Actually CAN Search Your ChatGPT Conversation History 🔍
ChatGPT is missing a major feature that makes finding old conversations a hassle: no search function for your chat history on desktop.
With multiple conversations spanning many topics, we've all struggled to dig up a specific chat from weeks or months ago. Endless scrolling seems like the only option.
But here's a little-known hack: ChatGPT actually does have a search feature, but ONLY in its mobile app.
It’s located right on top above the ChatGPT logo (see image 👇).
Just tap the search icon and type in keywords to instantly find any conversation.
Once I find the desired conversation on mobile, I enter any text into the prompt box to pull that conversation to the top of my conversation list on desktop.
It’s that easy, if a bit annoying. But also SUPER useful.
FAVORITE FINDS
What if you could transform your daily walk or commute into a productive reading session? ✨
Meet my new obsession: The ElevenLabs Reader app.
It allows you to listen to any article, PDF, ePub, document or text on the go with the highest quality AI voices. Note: it won't read with links from paywalled sites.
You can even paste in ANY content including article links to have it read out loud.
Plus, it’s super easy to use and you get 3 months of free access with unlimited generations.
This app has transformed my morning routine: Instead of scrolling through articles over coffee while sitting, I now grab my coffee and go for a 30-minute walk while listening to them. So, I can get my steps in while starting my day on a more energetic note.
You can download the iOS app here and learn more about the app here.
If you start using it, let me know how you like it.
That's all for this week.
I’ll see you next Friday. Thoughts, feedback and questions are welcome and much appreciated. Shoot me a note at avi@joinsavvyavi.com.
Stay curious,
Avi