Nvidia's big slip 📉, AI's secret deal 🤝, and Google’s controversial AI move 👀

Welcome back to a new edition of AI Odyssey!

TL;DR

• Nvidia’s stock dip: Despite a 122% revenue jump, Nvidia’s stock dropped 8% due to unmet investor expectations and a cautious future outlook.

• AI firms to share models: OpenAI and Anthropic will give the U.S. government early access to their AI models, allowing for risk assessment before widespread deployment.

• California AI bill divides tech: California’s proposed AI bill, SB 1047, could set a precedent for U.S. regulation, sparking a debate between innovation and safety.

• Google updates AI image tool: Google’s Gemini AI will resume generating images of people with new safety measures to prevent inaccuracies and misuse.

^{Sponsored by Synthesia}

Scale your in-house video production with AI.

Say goodbye to the bottlenecks of traditional video production. With Synthesia Enterprise, you can create stunning, on-brand videos faster than ever before, all powered by cutting-edge AI.

Whether it’s for training, marketing, or sales, our platform scales effortlessly to meet your needs. Collaborate securely with your team, tap into 230+ AI avatars, and deliver content in 140+ languages—all while keeping your data safe with enterprise-grade security and ethical AI practices.

Join 50,000+ companies already transforming their video strategy.

👉 Book your demo today!

1: Nvidia’s stock dips despite record revenue boom

Nvidia reported a staggering 122% rise in revenue, hitting $30 billion last quarter. But even this massive growth couldn’t meet Wall Street’s sky-high expectations, causing shares to drop 8% after hours.

Why it matters:

Nvidia’s chips are central to the AI boom, powering tech giants like Google and Microsoft. While its stock has skyrocketed 160% this year, investor expectations have soared too, leading to disappointment with a cautious forecast of $32.5 billion for the next quarter.

The bottom line:

Nvidia’s record growth underscores its role in the AI surge, but soaring market expectations mean even standout performance might not be enough to satisfy investors.

2: AI giants pledge to share new models with U.S. government

OpenAI and Anthropic will grant the U.S. government early access to their latest AI models, as announced Thursday.

Why it matters:

As AI technology advances, governments are pushing for more oversight to assess the risks of new algorithms. This move allows the U.S. to evaluate AI models before they're widely deployed.

Driving the news:

The companies signed agreements with the U.S. Artificial Intelligence Safety Institute, part of the Commerce Department, to collaborate on assessing AI safety and developing risk mitigation strategies.

The bottom line:

This partnership marks a significant step in setting standards for responsible AI development and may influence global approaches to AI regulation.

3: California’s controversial AI bill divides tech world

California's proposed AI regulation, SB 1047, is creating a rift in the tech community, with some seeing it as necessary oversight and others as overreach by the state.

Why it matters:

As Europe pushes ahead with AI regulations, the U.S. remains in legislative limbo, leaving states like California to take the lead. SB 1047 could set a precedent for how AI is regulated in America.

The details:

SB 1047 would require AI developers to meet specific safety standards before releasing their models. The bill passed the California Senate with overwhelming support and faces an August 31 deadline for Assembly approval.

Driving the debate:

Anthropic supports the bill with revisions, while OpenAI argues it could hinder innovation. Elon Musk has endorsed the bill, citing AI’s potential risks to public safety, while critics like Nancy Pelosi and tech groups urge a veto, calling the bill flawed and misdirected.

Key changes:

The bill has been revised following industry feedback:

• No criminal penalties; only civil penalties remain.

• Civil penalties can only be sought after harm has occurred.

• Reduced regulatory burden for open-source developers spending less than $10 million on existing models.

The bottom line:

With Congress slow to act, California is pushing forward to shape AI policy, potentially setting a new standard for tech regulation in the U.S.

4: Google updates Gemini's AI image creator to resume generating people

Google announced Wednesday that it’s updating its Gemini AI image creator with its latest generator, Imagen 3, and will soon resume the creation of images featuring people.

Why it matters:

Earlier this year, Google halted its AI-generated images of people after reports surfaced of historically inaccurate depictions, like Black founding fathers. The new update aims to address these issues while allowing for more creative flexibility.

Driving the news:

Google’s blog post outlines that the updated image generator will roll out to Gemini users in the coming days, with specific restrictions to enhance safety:

• No photorealistic images of people.

• Excludes images of children and identifiable individuals.

• Prevents depictions of violent, gory, or sexual scenes.

Currently, these features are only available for English prompts and are restricted to Gemini Advanced, Business, and Enterprise users.

Additional updates:

Google is also expanding its "Gems" feature, which allows for domain-specific versions of Gemini. Initially previewed at Google I/O, Gems are now available in 150 languages for paid users within the Gemini platform.

That’s a wrap! See you again for a fresh dose of AI Odyssey soon. 😎