HARDWARE

OpenAI adds Broadcom to deal stockpile

OpenAI added another chip deal to its list. 

The tech giant announced a deal with chipmaker Broadcom on Monday to develop and deploy 10 gigawatts of custom AI chips and systems over the next four years. The exact financial terms of the agreement weren’t disclosed, but sources told The Wall Street Journal that the deal is worth multiple billions of dollars. Broadcom’s share price jumped 9% following the news. 

OpenAI is also reportedly working with Arm to develop a CPU that works in tandem with Broadcom’s technology, according to The Information. 

The partnership adds yet another to the waterfall of deals OpenAI has inked in recent weeks. 

• In September, it forged a partnership with Nvidia to build at least 10 gigawatts of AI data centers and announced further buildouts as part of the Stargate project, bringing the initiative to 7 gigawatts and $400 billion in investment. 

• It also bumped up its contract with CoreWeave by $6.5 billion to deploy AI chips, bringing its total contracts with the firm to $22.4 billion, and announced a partnership with AMD last week to deploy 6 gigawatts of its GPUs.

On the OpenAI podcast on Monday, OpenAI CEO Sam Altman called the current push towards building out AI infrastructure “The biggest joint industrial project in human history.” Still, he noted, it’s “a drop in the bucket compared to where we need to go.”

This deal also adds another layer to the massive question of circular financing in the AI industry, with each of these companies continuing to prop one another up and add zeros to their billions in contracts.

Sam Altman and other OpenAI leaders have long held the view that AI is going to supercharge humanity. The availability of compute is one of the biggest factors in this success: OpenAI President Greg Brockman noted in Monday’s podcast that humanity will need “as much compute as possible” to achieve breakthroughs that wouldn’t otherwise be possible, and that demand is already far outstripping availability. 

But with great power comes great responsibility: If this lofty vision does come to pass, the organizations at the center of it stand to hold a lot of power and the ability to shape the future of technological innovation as a whole. And since compute is as good as gold, in inking these massive deals, OpenAI might be seeking to position itself as the proprietor of that power – both figuratively and literally.

TOGETHER WITH IBM

3 actions to help boost agentic AI impact

Businesses face many challenges going from AI ambition to implementation.

Success requires more than adopting new technology. It involves intentionally aligning AI with workflows, data, and goals.

As AI evolves into a business-wide opportunity, leaders must ensure the right AI is applied to the right challenges.

Explore IBM’s playbook detailing 3 challenges and approaches you can take to help boost the impact of your agentic AI.

• Finding your AI problem

• Creating a clear AI plan

• Integrating AI with the tools you use every day

Read the playbook

POLICY

California cracks down on chatbots, social media

On Monday, California Governor Gavin Newsom signed a new wave of laws regulating the way that tech platforms interact with young people. 

Newsom signed into law SB 243, a bill regulating AI companions as a means of protecting young users, making it the first state to do so. The law, which goes into effect on January 1, 2025, requires chatbot developers to incorporate features such as age verification and warnings and establish protocols to address discussions of self-harm. 

Additionally, Newsom signed AB 1043, requiring device makers and app stores to verify users’ ages online, a law that was backed by the likes of Google, Meta and OpenAI, and AB 56, which requires health warning labels for social media platforms for “profound” mental health risks.

“Emerging technology like chatbots and social media can inspire, educate, and connect – but without real guardrails, technology can also exploit, mislead, and endanger our kids," Newsom said in a statement on Monday. 

Newsom noted in his statement that a number of “horrific and tragic” cases involving AI chatbots have resulted in harm to young people, such as the allegations that ChatGPT is responsible for the death of a 16-year-old boy. A month after the lawsuit was filed, OpenAI announced a suite of parental controls that automatically limit certain content and provide warnings for acute distress, as well as an upcoming age prediction system. 

More teens than ever are turning to AI for companionship and comfort without the friction of human relationships, with one study showing that one in five high schoolers have had or know someone who has a romantic relationship with an AI model. 

As many of these platforms seek to reach young audiences and expand their AI capabilities, such laws may become even more relevant.

TOGETHER WITH AUTH0

Want To Connect Your AI Agents to Apps and APIs?

Deep View readers can do this using Token Vault, an Auth0 for AI Agents feature.

Auth0 for AI Agents addresses one of the biggest problems facing the AI landscape today: security. When you’re dealing with sensitive information and critical data, it’s crucial that you can trust your agents completely – and with this new solution, Auth0 lets you do just that. Token Vault lets you integrate your apps and agents with third-party tools, so they can do everything from handling and securely storing access tokens to connecting users to external tools, processing authorization requests, and more.

What does that mean for builders like you? The ability to stand up secure, trustworthy AI agents with just a few lines of code. Not too shabby, when you put it that way. Ready to check it out? Get started with Token Vault today.

MARKETS

JPMorgan Chase fuels AI efforts

Financial institution JPMorgan Chase announced an investment of up to $10 billion in companies and industries it deems “critical to national security.”

The decade-long investments will focus on four key areas: “frontier” technologies like AI, cybersecurity and quantum computing, energy dependence, defense and aerospace and supply chain and manufacturing. 

“This new initiative includes efforts like ensuring reliable access to life-saving medicines and critical minerals, defending our nation, building energy systems to meet AI-driven demand and advancing technologies like semiconductors and data centers,” CEO Jamie Dimon said in the firm’s announcement.

Although this investment is spread among several sectors, JPMorgan Chase has long been one of the most AI-forward financial institutions on Wall Street. 

• This effort began in earnest last fall, when it announced a mandate that every new hire needed to be trained on AI and provided all of its employees with access to its AI assistant called LLM Suite. 

• Dimon also noted last year that the company would add thousands of AI-focused jobs in the next several years. 

• The company is now reportedly deploying agentic AI to handle complex tasks on behalf of employees. 

And this focus appears to be working in the firm’s favor. Dimon told Bloomberg last week that the company spends $2 billion annually on AI in areas such as risk, marketing and customer service, but that the cost savings enabled by the tech now allow it to break even.

LINKS

• Anduril to deliver 100 units of AI-powered helmet to U.S. army personnel 

• Self-driving company Wayve in talks with SoftBank, Microsoft for $2 billion investment

• Oracle CEO Clay Magouyrk says OpenAI can cover $60 billion annually in cloud use

• AI insurance industry ClaimSorted raises $13.3 million in seed funding

• Meta is bringing back job boards on Facebook for entry-level work

• Nanochat: A full-stack training and inference pipeline of an LLM into a single “dependency-minimal codebase.”

• Microsoft MAI-Image-1: Microsoft’s first text-to-image generator developed in house, boasting photorealistic imagery.  

• Agentforce 360: The latest version of Salesforce’s agentic platform, including new ways to instruct and deploy agents, as well as Slack integrations. 

• Plane: AI-powered project management platform that centralizes the entire organization into one workspace. 

• Sigmamind.ai: A conversational AI platform that lets you build and deploy enterprise-grade agents rapidly.

• Mode Mobile: 32,481% 3-year growth. Deloitte’s #1 software company. Accredited investors get in early at $0.50/share. (sponsored)

• Nvidia: Research Scientist, Trustworthy AI

• Datadog: Senior Developer Advocate, Generative AI 

• OpenAI: Protection Scientist Engineer, Intelligence and Investigations

• Palo Alto Networks: Staff AI Scientist

GAMES

A QUICK POLL BEFORE YOU GO