Last year, researchers from Google and Stanford University populated a virtual village with about two-dozen ChatGPT-powered characters. The Sims-like personas could remember their roles, go on dates, challenge each other to competitions; they even coordinated to throw a Valentine’s Day party.

Believe it or not, this sleepy little AI town held some clues about a new stage in the generative AI hype cycle. Tech giants and startups alike are increasingly thinking about how large language models (LLMs) can move beyond chatbots and into autonomous “agents” that can perform tasks on their own accord. More than just answering questions and proffering information, this new crop of systems tap LLMs to actually complete multi-step actions, from developing software to booking flights.

While the tech is still relatively nascent, it’s progressed in the past year. Google DeepMind recently unveiled an AI agent called SIMA that was trained on 3D video games, including something called Goat Simulator 3, to handle around 600 skills, from navigating the games to opening menus, demonstrating what the research arm called “the potential to develop a new wave of generalist, language-driven AI agents.”

Mike Gozzo, chief product officer at customer service AI company Ada, said he’s seen a shift from a nearly singular industry focus on retrieval-augmented generation—LLMs that can navigate and interact with big databases of information—to autonomous agents in recent months. He said Ada has been working with agents since around the time the tech first began to pique developer interest last year, when OpenAI rolled out Auto-GPT.

“I’ve got this meeting with [OpenAI CEO] Sam [Altman] coming up later today. I was in DC with [Microsoft CEO] Satya [Nadella] a few weeks ago. And like, that’s all anyone is talking about is how do we get to autonomy in these systems?” Gozzo said. “And the reason why it’s even relevant at all is that, far beyond customer service, I think there’s going to be a trend in all software to move through this autonomous agent-style workflow.”

Keep reading here.—PK

Did you know it takes 10k iPhone batteries’ worth of lithium to make one EV? With over 350m EVs projected to be sold globally by 2030, demand for lithium will soar. Current extraction methods are too slow and inefficient to keep up.

So when EnergyX revealed that their technology could extract 300% more lithium than traditional methods, investors everywhere took note.

They’ve inked deals with top lithium producers like POSCO (who’s building a 100k ton/year plant), secured a $50m strategic investment from GM, and won a $5m DOE grant. They’ve even acquired a 90k-acre mining asset with ~5.5m tons of lithium.

Now, EnergyX is accepting shareholders for a limited time. Become an EnergyX shareholder here.

Meta is pouring a lot of money into AI right now, and it could be a long while before those bets start to bear fruit.

That was the message CEO Mark Zuckerberg delivered at the top of Meta’s earnings report last week, sending investors scurrying to sell off the company’s stock. Shares of the social media giant dipped as much as 17% in after-hours trading on Wednesday, though the price rebounded slightly the next day.

Never mind that the company delivered revenue and earnings that actually narrowly outperformed analyst expectations. Investors were likely spooked by a forecast that included more spending on AI projects and slowing revenue next quarter.

Zuckerberg claimed any stock turbulence is to be expected, noting Wall Street’s reactions to the company’s early spending on other big projects like Reels, Stories, and the transition of the News Feed to mobile.

“It’s worth calling out that we’ve historically seen a lot of volatility in our stock during this phase of our product playbook, where we’re investing in scaling new products but aren’t yet monetizing,” Zuckerberg said on the call.

Keep reading here.—PK

Coworking is a weekly segment where we spotlight Tech Brew readers who work with emerging technologies.

How would you describe your job to someone who doesn’t work in tech?

There’s an expression that I like to use: It’s a bit like “walking in the valley of death while chewing on glass.” That gets at the gritty reality and the thrill of what we do at Mendel AI. On a more practical level, my job is finding ways to program machines to approximate the intricate thought processes of a doctor’s brain at scale. To achieve our goals, I surround myself with incredibly smart data scientists and clinical experts.

Our team, which is primarily based in the US, is dedicated to pushing the boundaries of what’s possible in healthcare AI. My job ‌is about fostering a culture of collaboration where every team member, regardless of their role, feels integral to the process. It’s about ensuring they see the value and purpose in their work, especially after months of developing and tweaking models. When these models finally start working, it’s a testament to the collective effort and a moment of validation for everyone involved.

What’s the most compelling tech project you’ve worked on, and why?

The most compelling tech project I’ve worked on, without question, is Mendel AI.

We live in an era where we’re swimming in data but starving for information, as vast amounts of clinical data exist, yet extracting actionable insights remains a formidable challenge. This paradox was the catalyst for Mendel AI.

Keep reading here.

Morning Brew is partnering with Scott Galloway’s Section to bring you the AI Crash Course, a 10-day deep dive into AI. You’ll learn to write effective prompts, set up your “AI workspace,” audit your workflows, and start using AI for basic use cases. You’ll also get dedicated office hours with an AI expert and a suite of AI resources you can take with you. Enroll now and get 20% off with our exclusive code: BREW20.