Plus AI Outperforms Humans (subscribe in the links below)

Get a free 20-page AI explainer: AI FROM ZERO plus these stories and more, delivered to your inbox, every weekday. Subscribe to our newsletter at https://aidaily.us

Like this? Get AIDAILY, delivered to your inbox, every weekday. Subscribe to our newsletter at https://aidaily.us

OpenAI in Talks with Scarlett Johansson Over AI Voice Similarity

OpenAI is in discussions with Scarlett Johansson's team after pulling a ChatGPT voice, "Sky," that resembled her voice from the film Her. The company did not intend to mimic Johansson and is taking feedback seriously. OpenAI emphasizes customization to make user interactions feel natural and comfortable.

AI Learns to Draw Inspiration, Not Copy

Researchers at The University of Texas at Austin have developed Ambient Diffusion, a framework to train AI models using corrupted images to avoid replicating original works. The technique maintains high-quality image generation while reducing memorization of source data. This method, beneficial for scientific and medical applications, addresses ethical concerns in AI image creation.

AI Models Outperform Humans in Mental State Tests

New research in Nature Human Behavior shows AI models, like OpenAI’s GPT-4 and Meta’s Llama, match or surpass humans in tests for understanding mental states, known as “theory of mind.” While these models excel in tasks involving irony and indirect requests, they don't truly grasp human emotions. Experts caution against overestimating AI's human-like abilities.

New AI Algorithm Enhances Autoimmune Disease Prediction and Treatment

A new AI algorithm developed by Penn State researchers improves predictions and therapies for autoimmune diseases by analyzing genetic data more accurately. Named EXPRESSO, it identifies genes linked to autoimmune conditions, such as lupus and Crohn's disease, and suggests potential treatments. This advancement offers hope for earlier interventions and better therapies.

Windows AI Feature Raises Privacy Concerns

Microsoft's new AI-powered "Recall" feature for Windows 11 records users' activities by taking snapshots of the screen every few seconds, storing them locally and encrypted. While designed to help users retrieve past activities, it raises privacy concerns. Users can manage or delete captured content, but the feature requires specific hardware and ample storage.

AI Boosts Creativity and Productivity in Marketing

AI is transforming industries, with 75% of marketing teams considering it essential. Canva's AI tools enhance creativity and efficiency, allowing employees to generate on-brand assets quickly. AI aids in design, writing, and brainstorming, freeing up creative teams for strategic work. However, data privacy and security concerns remain. As AI adoption grows, companies must build guardrails around its use.

Plus AI Pessimists Miss Opportunities (subscribe in the links below)

Get a free 20-page AI explainer: AI FROM ZERO plus these stories and more, delivered to your inbox, every weekday. Subscribe to our newsletter at https://aidaily.us

Like this? Get AIDAILY, delivered to your inbox, every weekday. Subscribe to our newsletter at https://aidaily.us

Could AI Companions Help Alleviate Loneliness?

In the age of AI, virtual companionship offers new ways to combat loneliness. AI-powered platforms provide personalized, empathetic interactions, catering to emotional needs and offering support. OpenAI CEO Sam Altman predicts that AI will enhance human connection, potentially revolutionizing how we interact and connect with each other in the future

AI Pessimists Overlook the Opportunities for Progress

Many fear the disruptions generative AI might cause, but they often overlook the potential benefits. AI can enhance education by acting as a personal tutor for each student, improving overall learning outcomes. AI can also counteract the spread of misinformation through tools like watermarking. While AI might replace some jobs, it will create new opportunities, just as past technological advances have done. The focus should be on how AI can address challenges and foster economic growth, rather than only on the risks

Google's AI Shift Could Speed Up Web's Decline

Google's move to AI-generated search results, replacing traditional links, may make the web more convenient but less engaging. This change risks reducing traffic to original content creators, weakening the web's knowledge pool. With AI summarizing information, people might contribute less, affecting the quality of data that AI relies on

New AI Discovery Could Lead to True Machine Reasoning

MIT researchers introduced "libraries of abstraction" to help AI learn tasks similarly to humans. These libraries, tested on large language models, showed significant improvements in tasks like programming, planning, and robotic actions. This breakthrough may bring AI closer to human-like reasoning, showcasing the brain's remarkable power

New AI Learns to Detect Sarcasm

Researchers at the University of Groningen developed an AI-driven sarcasm detector trained on text, audio, and emotional content from sitcoms like "Friends" and "The Big Bang Theory." The AI correctly identified sarcasm 75% of the time, promising smoother human-machine communication and potential applications in detecting negative language and abuse

AI Startup LOVO Faces Lawsuit Over Voice Misappropriation

Actors Paul Skye Lehrman and Linnea Sage, along with others, filed a class action lawsuit against AI startup LOVO, accusing it of using their voices without consent. The suit, filed in New York federal court, alleges that LOVO misappropriated voices from A-list talent like Scarlett Johansson and Ariana Grande for its AI voice tool. This case highlights growing tensions between creators and AI firms using unlicensed data to develop their technology