Welcome to Daily Zaps, your regularly-scheduled dose of AI news ⚡️ 

Here’s what we got for ya today:

• What is California SB 1047?

• AI data center race

• OpenAI’s copyright dilemma

• Amazon Alexa + Anthropic’s Claude

Let’s get right into it!

Sign Up | Advertise | Tools

GOVERNMENT

What is California SB 1047?

California lawmakers have passed Senate Bill 1047, a controversial bill that requires companies investing significant amounts in developing or modifying AI to test their models for potential societal harm, such as cybersecurity threats or the creation of dangerous weapons. Despite opposition from major tech companies like Google, Meta, and OpenAI, which argue that the bill could stifle innovation and harm startups, the bill passed with strong support in both the Senate and Assembly.

Supporters, including some industry whistleblowers and AI experts, believe the regulation is necessary to ensure AI development is safe and responsible. The bill now awaits the decision of Governor Gavin Newsom, who must decide whether to sign it into law.

BIG TECH

AI data center race

The development of AI relies heavily on specialized data centers that house vast clusters of AI chips, primarily from Nvidia. These facilities, which are critical for training advanced AI models, are highly secretive and costly, with planned centers in the U.S. expected to exceed $50 billion in construction costs alone. Major players like Microsoft, OpenAI, Meta, and Elon Musk's xAI are in fierce competition to build the most powerful supercomputers, some housing up to 100,000 chips, requiring massive power supplies and sophisticated cooling systems.

The race to build these AI supercomputers is driven by the belief that greater computing power will enable AI to achieve superhuman capabilities, potentially solving global problems. However, this competition has sparked tensions over chip access and technological constraints, with companies like OpenAI and xAI vying for dominance in the rapidly evolving AI landscape.

STARTUPS

OpenAI’s copyright dilemma

OpenAI's significant advancements in its GPT model likely came from scraping vast amounts of content from the web, including material from major publishers like Axel Springer and The New York Times, without permission. Despite this, OpenAI has struck deals with these publishers, possibly as a way to avoid lawsuits and gain real-time access to their content, enhancing its AI capabilities.

However, these deals raise questions about the sustainability and ethics of using scraped content, as well as the future of search engines like Google's, which is also navigating similar legal and competitive challenges. If courts rule against OpenAI in copyright cases, it could face substantial financial penalties, potentially altering the AI landscape by favoring larger companies that can afford to pay for training data, while smaller competitors struggle.

BIG TECH

Amazon Alexa + Anthropic’s Claude

The new Alexa, which might include features like AI-generated news summaries and a chatbot for children, has struggled in early tests using Amazon's own AI models, leading to the possibility of incorporating third-party AI, such as Claude from Anthropic. 

Amazon's new generative AI-powered Alexa, expected to be more advanced and capable of handling multiple smart home tasks with simpler commands, may face delays and challenges before its release. Initially planned for June and now possibly October, this version could come with a subscription fee of $5 to $10 per month, a shift from the traditionally free Alexa service.

📮 Rate Daily Zaps