Good morning! This is your daily ☕️ Techpresso.
In today's Techpresso:
🎥 OpenAI announces new text-to-video platform, Sora
🍎 Apple is pushing generative AI efforts with new tools for developers
🤖 Meta's new AI model learns from videos rather than text
⚖️ FTC wants to penalize companies for use of AI in impersonation
🎁 + 6 other news you might like
🔮 + 3 handpicked research papers and tools
🎥 OpenAI announces new text-to-video platform, SoraLINK
OpenAI introduced a new generative video model called Sora, capable of converting short text descriptions into detailed, high-definition video clips up to a minute long.
Sora marks a significant advancement in text-to-video generation technology, displaying an ability to handle complex visual tasks like object occlusion and maintaining consistent styles across cuts.
Despite its potential, OpenAI has not yet released Sora to the public, focusing on safety tests and feedback from video creators to mitigate risks associated with deepfake technology.
🍎 Apple is pushing generative AI efforts with new tools for developersLINK
Apple is internally testing new generative AI features for its Xcode programming software, with plans to release these updates to third-party developers within the year.
The company is exploring generative AI for consumer applications such as creating playlists in Apple Music, generating slideshows in Keynote, and enhancing Spotlight search with chatbot-like capabilities for natural language queries.
Apple has recently released AI tools including the MLX machine learning framework, the MGIE text-to-image editing model, and the Keyframer AI animator, while also working on a programming tool similar to GitHub Copilot and Amazon’s CodeWhisperer.
🤖 Meta's new AI model learns from videos rather than textLINK
Meta has developed a new AI model, Video Joint Embedding Predictive Architecture (V-JEPA), that learns from videos instead of text, aiming for a more human-like learning process.
V-JEPA analyzes unlabeled videos to understand interactions and form an internal model of the world, a step towards achieving Artificial General Intelligence according to Meta's FAIR group head, Yann LeCun.
The model could revolutionize Meta's augmented reality glasses and AI training methods, with plans to further enhance V-JEPA by incorporating audio learning capabilities and releasing it under a Creative Commons noncommercial license.
⚖️ FTC wants to penalize companies for use of AI in impersonationLINK
The U.S. Federal Trade Commission (FTC) is proposing new rules to hold companies accountable if their technology, especially AI-generated content, is used for impersonation scams.
The FTC has finalized a rule targeting impersonations of businesses and the government, aiming to fight scams using false representations such as business logos or government addresses.
FTC Chair Lina Khan expressed concern over the surge in impersonation fraud, exacerbated by the use of AI tools capable of replicating individuals’ voices and other characteristics with high accuracy.
Other news you might like
OpenAI's Altman seeks US government nod for AI chips venture, Bloomberg reports.LINK
Apple makes it official: No Home Screen web apps in European Union.LINK
First human trial confirms safe advancement of 'wonder' nanomaterial.LINK
EU watchdogs are concerned about Meta's "Pay or Okay" model for ad-free subscriptions.LINK
After rockets and spacecraft, Rocket Lab’s next frontier could be applications.LINK
Microsoft is slowly building to a future where ‘every screen is an Xbox’.LINK
Latest research and tools
Hexapod project: a DIY guide on building your own 3D printed robot with videos, community support, and detailed parts lists.LINK
Orgzly Revived: a community-maintained project that continues the development of the Orgzly app for note-taking and to-do list management, supporting Org Mode's plain-text system on Android.LINK
Personality Trait Recognition Using ECG Spectrograms and Deep Learning: this study shows how ECG (heart activity) patterns combined with artificial intelligence can predict a person's personality traits.LINK
Want to get the latest news differently? Find us on:
See you tomorrow for a new dose of ☕️ Techpresso!
-min.png){kind=small}
.png){kind=small}
