Artificial Intelligence Newsletter

AI NEWSLETTER / 16.05.2023

Spotify expands AI-powered DJ feature to UK and Ireland

Spotify is launching its AI feature called “DJ” for premium customers in the U.K. and Ireland. The feature is aimed at music discovery and it plays curated music with AI-powered commentary in English. In many ways, using the AI DJ feels like listening to a radio DJ.

The company first launched?this feature back in February?for premium subscribers in the U.S. and Canada. The feature is powered by OpenAI’s tech and is still in beta, so there might be some glitches.

After the feature launch, the company noticed that Gen Z enjoyed the new DJ feature as they made up 87% of DJ users. Plus, the company found out that people who use the AI DJ on a particular day will spend 25% of their listening time with the feature.

[source]

Google I/O 2023 is a wrap — here’s a list of everything announced

On Google I/O keynote day, the search and internet advertising provider put forth a rapid-fire stream of announcements during its developer conference, including many unveilings of recent things it’s been working on.

• Google Maps: Google Maps unveiled a new “Immersive View for Routes” feature in select cities. The new feature brings all of the information that a user may need into one place, including details about traffic simulations, bike lanes, complex intersections, parking and more.?Read more

• Magic Editor and Magic Compose: Google’s Magic Editor feature is AI-powered for more complex edits in specific parts of the photos — for example, the foreground or background — and can fill in gaps in the photo or even reposition the subject for a better-framed shot.?Check it out

• PaLM 2: Frederic?has your look at PaLM 2, Google’s newest large language model (LLM). He writes, “PaLM 2 will power Google’s updated Bard chat tool, the company’s competitor to OpenAI’s ChatGPT, and function as the foundation model for most of the new AI features the company is announcing today.” PaLM 2 also now features improved support for writing and debugging code.?More here

• Bard gets smarter: Google is not only?removing its waitlist?for Bard and making it available, in English, in over 180 countries and territories, but it’s also launching support for Japanese and Korean with a goal of supporting 40 languages in the near future. Also new is Bard’s ability to surface images in its responses.?Find out more

• Workspace: Google’s Workspace suite is also getting the AI touch to make it smarter, with the addition of an automatic table (but not formula) generation in Sheets and image creation in Slides and Meet. Initially, the automatic table is simpler, though?Frederic?notes there is more to come with regard to using AI to create formulas. The new features for Slides and Meet include the ability to type in what kind of visualization you are looking for, and the AI will create that image. Specifically for Google Meet, that means custom backgrounds.?Check out more

...

[Read More of the List]

EU nears ban on predictive policing and facial recognition after AI Act vote

The EU is edging closer to a landmark ban on predictive policing and?facial recognition.

At a crunch vote in the last week on the bloc’s flagship?AI?Act, two committees of MEPs overwhelmingly endorsed sweeping new rules on artificial intelligence. The text now moves to a vote by the entire European Parliament in June.?Once approved, the regulation will become the world’s first comprehensive AI law.

At Thursday’s vote, MEPs approved a?strengthened version of the rulebook. The Act now prohibits predictive policing and facial recognition in public spaces.

The amendments also introduce new restrictions on generative models, such as ChatGPT, and emotion recognition.

The industry responds

In the tech sector, reactions to the vote were mixed.?The Software Alliance (BSA), a lobby group that represents the likes of Microsoft and IBM, has called for further clarification.

“The enterprise software industry remains concerned about the allocation of responsibilities in the?AI?value chain and the treatment of foundation models,” said Matteo Quattrocchi, BSA’s policy director.

[source]

ChatGPT creator OpenAI is releasing an open-source AI model: Report

An open-source artificial intelligence (AI) model is reportedly being prepared for public release by OpenAI, the firm behind the AI chatbot ChatGPT.

In a May 16?report?in The Information citing a person with knowledge of the plan, OpenAI is undertaking the move as pressure mounts from competing open-sourced AI models, such as those leaked from Meta in February.

The timeline of when the model would be released was not reported.

It was said OpenAI’s open-source model would likely not be competitive with its flagship ChatGPT product, as the firm’s value comes from being able to sell access to its more sophisticated models.

OpenAI has faced stiff competition from open-source AI models such as Meta's LLaMa — which was originally limited to researchers but was leaked in full by a user from the imageboard site 4chan in late February.

Other open-source models include those from Stability AI, which?opened its large language models?in April, along with Databricks’ Dolly 2.0 AI, which it open-sourced?days?prior to Stability AI.

[source]

You can subscribe to our?Artificial Intelligence e-mail newsletter?and get weekly news from your e-mail.

Also, check out our?data annotation management dashboard?whenever you need data labeling!?

Boost the performance of your AI with high-speed, high-quality data annotations with?Co-one!