Zuckerberg's Vision for AI: A Bot That Makes Ads and Helps You Say Happy Birthday to Your Friends

In a new podcast episode with Lex Fridman, Meta CEO Mark Zuckerberg discussed his upcoming plans with AI, including an AI assistant that can help your social life and make ads for your business.

Zuckerberg told Fridman that, unlike OpenAI's GPT models, he hopes to deploy a huge, diverse range of Large Language Models (LLMs) that are fine-tuned to different needs and businesses. One of these needs, Zuckerberg said, is companionship.

Most people probably don't have as many friends as they'd like to have. You'll get tools that help people connect with each other, [which] are gonna be the primary thing that we do," Zuckerberg said. You can imagine AI assistants that just do a better job of reminding you when it's your friend's birthday and how you can celebrate them. Having something that's more of a social assistant in that sense, that can update you on what's going on in their life and how you can reach out to them effectively and how you can be a better friend."

Zuckerberg said that the AI assistants he plans on launching will take on roles including a mentor, a life coach, a cheerleader that can help pick you up through all of life's challenges." Zuckerberg thinks AI is a form of social technology that can help people express themselves better to people in situations where they would otherwise have a hard time doing that." According to him, language is an imperfect way for people to express their ideas," so AI language models can help improve that barrier for Meta users.

He brings up an example of his newborn and five-year-old daughter who cry because they have a hard time expressing their thoughts and compares this to a lot of adults. A lot of adults get frustrated too because they have trouble expressing things," Zuckerberg told Fridman. All these different technologies that can help us navigate the social complexity and actually be able to better express what we're feeling and thinking, I think that's generally all good."

This assistance translates to business deals, according to Zuckerberg, who says creators will now be able to negotiate with brands using an AI assistant. [Creators] go into their DMs and see brands want to do something with them and they don't know how to negotiate or push back respectfully, so I think building a tool that can actually allow them to do that well is one simple thing that we've heard from a lot of people that they'd be interested in," he said.

Zuckerberg also suggested that through Meta's combined language and image models brands won't need to outsource ad agencies anymore to create commercials. In the future, if you're advertising on our services, do you need an ad creative? No, you just need to tell us, Okay I'm a dog walker and I'm willing to walk people's dogs and help me find the right people,' and [we'll] create the ad unit that will perform the best. Give an objective to the system and it connects you to the right people," he said.

Currently, Meta is one of the major players in the fight to develop the newest AI technologies, alongside companies such as Microsoft, which has partnered with OpenAI to integrate AI across its products, and Google, which has launched its own language model Bard and begun to deploy AI assistants across its workspace applications.

Meta's LLaMA language model was announced in February and was released to researchers, but was leaked a week later on 4chan and quickly spread across online communities. This launched a debate regarding the possibility of people misusing the model, while others argued that an open-source model is important so that people can transparently test and send feedback to Meta about the model.

In the podcast interview, Zuckerberg said that Meta's approach is to be open and academic in its development of AI. He said that he believes open-sourcing technology helps unlock a lot of innovation in the ecosystem and valuable feedback" and that he values community-driven training." This stance is a clear divergence from OpenAI's recent actions, which have been recently criticized for their secretive nature that have failed to disclose how its models are trained and what dataset was used to train them.

Zuckerberg also criticized the concerns surrounding AI's existential risk, including that of AI researcher Eliezer Yudkowsky who called for AI development to be shut down.

I've spent quite a lot of time thinking about it from that perspective. Over time, I think that we need to think about this even more as we approach something that could be closer to superintelligence. I think that it's just pretty clear to anyone working on these projects today that we're not there," he said. There are concerns that already exist about people using AI tools to do harmful things of the type that we're already aware. That's going to be a pretty big set of challenges that the company's worked on that we're going to need to grapple with, regardless of whether there is an existential crisis as well sometime down the road."

I do worry that to some degree, people can get a little too focused on some of the tail risk and then not do as good of a job as we need to on the things that we can be almost certain are going to come down the pipe as real risks that manifest themselves in the near term," he added.