Cloud Hosting

In a recent panel hosted by The Information in partnership with Comcast NBCUniversal LIFT Labs, several experts took a look at where artificial intelligence is now and where its going. Amir Efrati, executive editor at the Information, acted as moderator. The panelists included AI experts from the worlds of customer support, technology, entertainment and investing:

Isaiah Jenkins is senior manager of startup programming and employee engagement for Comcast NBCUniversal LIFT Labs.

Deon Nicholas is CEO and co-founder of Forethought, which produces generative AI for customer support.

Jadyn Bryden is a vice president of XFund, an early-stage venture capital firm focused on forming partnerships between venture capital firms and research universities.

Sara Hooker leads Cohere For AI, a nonprofit research lab that solves complex machine-learning problems.

The Most Exciting Uses for AI Now and in the Near Future

The panelists first tackled a scenario thats literally life and death: how AI can help speed medical diagnoses and make them more accurate. Sara Hooker laid out the scenario where medically trained AI can be most helpfulin remote areas that dont have access to doctors or hospitals. But she noted that AI-generated diagnoses also come with risks of inaccuracy.

The antidote, said Deon Nicholas, is creating more-specialized data sets. Sometimes models like chatbots can confidently spew nonsense. You dont want somebody getting the wrong diagnosis. Anytime lives or livelihoods are at stake, were going to need more of these next-generation proprietary models built on these specific data sets because theyre going to be more correct.

How Can Startups Get In on the AI Action?

Beyond medical applications, theres another group deeply interested in generative AI: startups. While a lot of new tech companies are eager to capitalize on the AI gold rush, what will make them successful? As with any other business, the key is differentiation. And whats different for AI now is producing new, proprietary data sets that enhance already available information. Bryden said shes also interested in companies that improve and organize current platforms. At a time like this, the startups that are really going to succeed are the ones doing what we call a picks and shovels play. Instead of trying to pick out whos going to be the company that strikes gold, you want to invest in the companies that enable the underlying levers of established AI platforms.

Pushing the Limits of Science

Coding and science are two other areas where AI has great potential. As Nicholas mentioned, now a developer can write code that talks back and helps them figure out what the next iteration of code could be. But AI may have even greater implications for science. Nicholas showed excitement about AIs ability to help us increase our understanding of the world. For the first time ever, people have the ability to query AI to help them with that research process. Lets say you wanted to push the boundaries of, say, quantum physics. What if you could ingest all of the dilemmas and theories of quantum physics? You wont necessarily have AI figure it out for you, but you could ask it, OK, what do I need to think about next? That is a shift forward for humanity.

AI as the Creative Muse

AI can not only inform but inspire. While the next Hollywood blockbuster probably wont rely on it, AI might very well offer prompts for scripts. Nicholas says writers are already using it as a creative muse. Theres this really interesting shift in how were going to do creative work with the power of generative AI. It can generate content. It can help you with the creative process.

Has AI Gone Too Far?

With the release of GPT4, natural language took such a huge leap forward that some people thought it warranted a pause in development. Two thousand expertsmany of whom had business interests that competed with ChatGPTcited the profound risks to society and humanity it posed. But is this fear warranted? And more importantly, has the train already left the station? Nicholas said, Its hard to put the genie back in the bottle. And I argue that even the mere fact that were having this much conversation about it is a good thing because it gives the public time to understand whats going on and then react to it. Bryden agreed: Its simply impossible to slow this down. People are excited about it.

Did you miss the panel? You cancatch the full replay on our site.

Follow this link:

Event Recap: What's Next for Generative AI - The Information