Recap: What’s Ahead for Enterprise AI in 2024

In 2023, companies all over the world rushed to take advantage of artificial intelligence. Now, these same companies are taking a breath and looking at their budgets. At the same time, we’re seeing incredible new use cases every week. In partnership with Comcast NBCUniversal LIFT Labs, The Information’s Amir Efrati discussed the future of AI with three experts in the fields of coding, security and large language models:

• Grant Bourzikas, chief security officer, Cloudflare 
• Lydia Ding, co-founder, CodeComplete AI
• Will Grannis, chief technology officer, Google Cloud 

Creating Better Customer Experiences With AI

Efrati opened up the conversation by questioning what advancements might be coming up for AI and asking the panelists to highlight interesting use cases.

Will Grannis of Google Cloud pointed out Gymshark as a great use case. The fitness retailer partnered with Google Cloud to help it during the busy holiday season. “They created a tailored shopping assistant for online shoppers. AI facilitates a customer getting product recommendations more quickly and expertly than trying to do this on your own.”

Can AI Beat Back the Hackers?

Corporations invest a huge amount of money and time on security, and for good reason. Grant Bourzikas, CSO of Cloudflare, described how AI can help fend off attacks more quickly. “A lot of times in security, we get wrapped up in the details because we have to spend time decoding or creating timelines of attacks. But AI allows me to see attackers that have similar patterns and offers a broader perspective because of the data involved in the models.”

Upping the Capacity for Cutting-Edge Coding

While AI has proliferated in almost every aspect of business, one domain it’s particularly useful in is coding. Efrati turned to Lydia Ding of CodeComplete to ask her about the future of AI in coding—or as he put it: “When does software begin to write itself?”

Ding went even deeper than coding. “AI has enabled us to reimagine how developers interact with code. And there are a lot of examples where a smaller, fine-tuned model can outperform the most popular generic models in specific scenarios. The smaller models can also be cheaper, more environmentally friendly and sustainable, which are all things that a business really needs to consider now.”

Is It Time to Curb the AI Spending Spree?

Only a few months ago, it seemed that companies were going all in with AI and spending money on the latest and greatest language models, or partnering with expensive vendors to create their own in-house models. But now businesses are considering costs more closely. Efrati asked the panelists what their outlook was in terms of spending: Are companies still going for the newest AI versions, or are slightly older versions capable of getting the job done just as well?

Grannis said, “Our customers want to have access to different models and platforms so they can choose their own innovation agenda in-house and not be limited by only having access to one model.” He added, “If you’re iterating over only the largest models, you know that that can be cost-prohibitive for a lot of very finite use cases and in midsize or smaller companies.”

What’s on the AI Horizon?

AI evolution isn’t slowing down anytime soon. So what’s next for AI?

While regulation can limit some innovation, Ding thinks it’s needed. “I think governance overall should be a big theme going forward. Companies will probably spend more time caring about those issues, as well as caring about the concrete output and not just chasing the next cool thing.”

Grannis was looking forward to seeing AI solve some of humanity’s most pressing problems. “There are roughly 400,000 stable materials that scientists were able to model that don’t exist in nature but could be created synthetically for power, solar energy or even semiconductors. These types of examples will create really exciting frontiers in the next few years.”

The Future of AI Is Wide Open

Even as some companies are approaching their tech spending more cautiously, AI is proving its mettle. It’s already shown that it can solve complex problems, reduce tedious tasks in the workplace and help humanity make immense scientific, software and health advancements in a fraction of the time. New frontiers indeed.