Welcome to this week's edition of Jumble, your go-to source for the latest in AI. This issue covers TikTok’s controversial new feature and how ChatGPT performs against human physicians. It’s a packed week in AI and here’s what you need to know ⬇️

TikTok Allows Content Creators to Clone Themselves With AI Avatars 👯‍♀️

TikTok is reportedly developing a feature that could revolutionize creator-fan interactions on the platform. The new project, codenamed "V," aims to enable creators to generate AI avatars that can engage with fans around the clock, even when the creators themselves are offline.

This concept, already being tested on Douyin (TikTok's Chinese counterpart), allows creators to build virtual representations of themselves based on their personality, thinking patterns, and previous content. These AI avatars would be capable of responding to comments, participating in live stream chats, and even handling direct messages, all while mimicking the creator's unique style and voice.

Potential Impact on Creator Economy

• 24/7 Availability: Creators could maintain a constant presence on the platform without burnout.

• Scalable Interactions: AI avatars could handle a higher volume of fan interactions than humanly possible.

• New Revenue Streams: This technology might open up new monetization opportunities for creators.

TikTok's exploration of AI avatars aligns with industry trends towards AI-enhanced social experiences. This move mirrors recent developments by other tech giants, notably Meta (formerly Facebook), which showcased similar technology at its Connect event. Meta's demonstration featured video avatars capable of interacting on behalf of creators. China has already seen success with virtual influencers, with over 993,000 digital avatar companies registered in the country.

Potential Challenges and Concerns

While the technology shows promise, it also raises questions:

• Authenticity: Will fans accept interactions with AI avatars as genuine engagement?

• Cultural Differences: The success of virtual influencers in China may not necessarily translate to other markets.

• User Experience: Can AI-driven interactions replicate the appeal of genuine human connection that social media platforms traditionally offer?

As TikTok continues to innovate, the "V" project could represent a significant shift in how creators and fans interact on social media platforms. However, its success will likely depend on user acceptance and the ability of AI to convincingly emulate human-like interactions.

Fully Automated Email Outreach With AI Agent Frank

Agent Frank was designed to fully take care of prospecting, emailing and booking meetings for you, so you can focus on closing deals!

With Agent Frank you won’t need 20+ different tools anymore - just set up your Agent in 4 steps and get started.

Learn More

AI Outperforms Doctors in Medical Diagnosis Study 🩺

A recent study has revealed that ChatGPT-4 outperformed human physicians in diagnosing medical conditions based on case histories. The study found that the AI system scored an average of 90% accuracy in diagnosing and explaining medical conditions, while doctors using the chatbot achieved 76%, and those without it scored 74%.

The experiment, involving 50 doctors from major American hospital systems, tested participants on six complex case histories. Researchers were shocked by the results, which challenge the conventional wisdom about AI's role in medical diagnosis.

The study also revealed two significant issues:

• Doctors' reluctance to consider AI suggestions that contradicted their initial diagnoses.

• Many physicians' lack of proficiency in utilizing AI chatbots to their full potential.

These findings suggest that while AI has the potential to be a valuable "doctor extender" offering second opinions, there is still a considerable gap in integrating this technology effectively into medical practice.

Bests and Busts

Here's a look at this week's AI highlights and lowlights:

⭐ Best: Scientists Are Using AI To Improve Vegan Meat Alternatives
Stanford University researchers are combining AI machine learning with mechanical testing to enhance plant-based meat alternatives, aiming to bridge the texture gap between animal and plant-based products. The team's innovative approach involves three-dimensional mechanical tests and AI data processing to analyze food textures, potentially accelerating the development of more realistic vegan meat options.

💩 Bust: Expert Believes AI Is Likely A Factor In Marriott Slashing Jobs
A business expert suggests that Marriott's recent decision to lay off 833 corporate employees may be linked to advancements in AI. Roland Rust, a professor at the University of Maryland's School of Business, notes that AI can now efficiently handle many corporate tasks such as data collection, analysis, and report generation, potentially reducing the need for human workers in these roles.

The Scoop 🍦

🧑‍🦳 AI Is Everywhere, and Boomers Don't Trust It
A new study reveals that Baby Boomers view AI tools like ChatGPT as a greater threat to data privacy than social media platforms, with 54% expressing this concern. The research also shows that 82% of Boomers are worried about the security and privacy of their personal data when using AI tools, highlighting a significant generational distrust in artificial intelligence technology.

🎙️ AI Voice Cloning Outpaces Legal Protections
Experts warn that AI voice cloning of celebrities is advancing faster than laws can keep up, with cases of fraud and unauthorized use on the rise. David Attenborough and other famous figures have fallen victim to voice cloning scams, highlighting the urgent need for updated privacy and copyright laws to address this growing challenge.

💼 How Secure Is The AI System Your Law Firm Is Using?
As law firms increasingly adopt AI for tasks like legal research and contract review, experts warn about the critical need to assess the security of these systems. The article emphasizes the importance of asking key questions about AI security, including data usage, retention policies, and compliance measures, to protect client confidentiality and maintain the integrity of the attorney-client relationship.

👓 Meta AI Begins Roll Out on Ray-Ban Meta Glasses in EU
Meta has announced the availability of its AI assistant, Meta AI, on Ray-Ban Meta smart glasses in France, Italy, Ireland, and Spain. Users in these countries can now interact with Meta AI using voice commands to ask general questions in English, French, Italian, and Spanish. However, the feature allowing AI to answer questions about what the user sees remains limited to the US, Canada, and Australia. This expansion marks Meta's efforts to comply with EU regulations while bringing AI functionality to more users, with plans for further expansion across the EU.

🛍️ Perplexity Launches AI-Powered Shopping Assistant 
Perplexity has introduced an AI shopping assistant called "Buy With Pro" for its US Pro users, enabling seamless product research and purchasing directly within the platform. The new feature offers one-click checkout for select products, free shipping, and a visual search tool called "Snap to Shop" that allows users to find items by uploading photos. This AI-driven shopping experience aims to streamline online shopping by providing unbiased product recommendations and easy comparisons.

🎵 They Taught AI to Sing, and It Was Beautiful
Artists Holly Herndon and Matt Dryhurst showcase their groundbreaking AI choral project "The Call" at London's Serpentine Galleries, featuring ethereal, AI-generated harmonies trained on diverse British choirs. While the auditory experience is described as uncanny and exquisite, the visual installations accompanying the exhibition are criticized for being cumbersome and detracting from the overall beauty of the AI music.

🔒 Microsoft Announces Its Own Hacking Event with Big Rewards 
Microsoft has launched Zero Day Quest, an in-person hacking event aimed at uncovering vulnerabilities in cloud and AI systems, offering $4 million in potential awards. This initiative, announced at the Ignite conference, builds on Microsoft's bug bounty program and provides security researchers direct access to Microsoft's AI engineers and Red Team, fostering collaboration to enhance security across its platforms. 

📚 HarperCollins Seeks Authors' Consent for AI Book Training
HarperCollins is offering authors the option to license their nonfiction books for AI training, with one author reporting a $2,500 offer for a three-year license. This move comes amid ongoing lawsuits against AI companies for alleged copyright infringement, as HarperCollins aims to balance innovation with protecting authors' rights and revenue streams.

Thank you for being a valued reader of Jumble! See you next week for more updates on the latest trends and developments in AI.

Stay informed, stay curious, and stay ahead with Jumble!

Zoe from Jumble