Rakesh Malhotra January 5, 2024
To say that 2023 was a transformative year in tech would be an understatement. Revolutionary, paradigm-shifting, disruptive, catalytic – it was all of these, and more. Now that the initial sprint to get on the AI leaderboard is over, the marathon begins for tech companies worldwide.
As we enter the second year of mainstream AI, the “new normal” of the intelligent era is starting to take shape. But what does that really mean in the near term?
Back in February 2023, I outlined three principles of AI with the caveat that “it is nearly impossible to predict where this wave of AI innovation will lead.” Nearly one year (and an incredible amount of learning) later, I no longer believe that’s true. So with that, here are my top four predictions for AI in 2024.
AI Prediction 1: Apple Puts an LLM in Your Pocket
ChatGPT made AI consumer-friendly – but not truly personalized. Even if your conversation with ChatGPT, Bard, or Claude feels personal, it’s still trained on an internet’s worth of content. It’s not yours and yours alone, and it’s not sized to run on your mobile devices. That’s a gap that the big players are already racing to fill (think Google’s Gemini Nano announcement in December 2023). But if we’re talking about world-class personal experiences and the tech that powers them, there’s a big name conspicuously missing from the list.
Apple has been largely silent with respect to AI being integrated into its flagship products – but that’ll change in 2024, when they put a private LLM in everyone’s pocket. Their unique integration across a vast ecosystem of products and services, ability to manufacture highly-performant and energy-efficient chips, non-reliance on ad revenue (and the resultant privacy bona fides) – not to mention their global customer base of 1.5B iPhone users – won’t just make Apple an AI force to be reckoned with. They will completely change the game.
AI Prediction 2: A Flood of AI-First Products
Since AI burst onto the scene, most companies have been testing the water by adding AI capabilities to existing products and services. The examples are everywhere: Google and Microsoft embedding AI into their search engines, Adobe infusing it into their design tools, and Miro enriching their collaboration tools, just to name a few. In 2024, companies will dive into the deep end: building products and services from the ground up, using AI technologies at their core. A new wave of AI-first innovators will introduce new search engines, design tools, and collaboration tools with AI as the primary and only interaction model for their users. The result? Redefined markets that lead to a wave of significant M&A activity not seen since the shift from mobile to cloud.
AI Prediction 3: Disrupting Professional Services for the Greater Good
Most professional services (think lawyers, accountants, IT) are billed based on units of time, in part because of the expensive, highly variable human cognitive labor they require. This approach has consequences; it forces firms to push the risk onto their clients, and creates a barrier to access that favors the wealthy. But what if you could bill for outcomes instead of hours?
In 2024, progressive firms will start using AI to mitigate much of that variability for a wide range of use cases, democratizing access to services by lowering the cost to clients. Counterintuitively, the use of AI to reduce variability will not translate to fewer humans providing these services; instead, expanded access will increase demand for those services and the humans that perform them (with the assistance of AI). In the coming decade, the boost in overall standards of living through availability of high-quality financial, legal, medical, and technological help would be akin to offering universal access to the internet, cell phones, or other consumer electronics.
AI Prediction 4: Multi-LLM Considerations Take Center Stage
As leaders grow more comfortable with AI in 2024, they’ll turn their attention from the “why” to the “what” and the “how.” The evolution of AI from concept to strategy will force organizations to grapple with practical aspects of adoption that may not have been on the radar last year. Here are a few notable ones.
There’s a lot of volatility and fast-paced change in the AI landscape right now, and that’s not likely to slow down in 2024; in fact, our bet is that it’ll increase. Organizations will need to be careful about how they roll out new AI-driven applications and services, ensuring developer and user experiences aren’t too tightly coupled with a specific LLM implementation. Even though standards in this space are pretty far away, leaders still need to put interoperability at the forefront – it will be crucial for functional, technical, and (last, but certainly not least) regulatory reasons.
Recently, our AI Practice CTO Alex Jettel shared his thoughts on how abstraction layers can play a role in solving this problem. IT leaders will need to consider approaches like this as they ramp up AI adoption in 2024.
Given the pace of innovation and the customizability of these models, it’s important that organizations test and use LLMs to understand which best fit their own needs. Not only is the space moving too quickly for organizations to wait for industry analyst firms to weigh in, but the effectiveness of the various LLMs will be highly dependent on the organization’s specific data sets and use cases. The only way to learn is by doing.
If you’re interested in experimenting with AI, but aren’t sure where to start, you may find inspiration in the insights we frequently share. Over the last several months, we’ve developed a proof of concept for an AI-powered business insight tool, introduced an open-source virtual agent accelerator for contact centers, launched a popular ChatGPT plugin to access video transcription data, created a custom GPT to streamline our content creation process, outlined a 6-category taxonomy for generative AI use cases, shared an in-depth assessment of popular LLM orchestration solution LangChain, and built an AI-powered cloud migration platform. We even developed a workshop for engineers who want to experiment with AI tools to boost their productivity. Make sure to watch this space for even more ideas and experiments in 2024.
Taking a Platform Approach
Last year, I posited that AI requires a platform approach. I believe that more firmly than ever as we venture into 2024. To keep up with an ever-evolving landscape of technologies and business needs, it’s critical to have a framework in which LLMs can be folded into the overall business strategy. The best way to do this? Think of your organization’s intelligence capabilities as being encapsulated by a platform.
By taking a platform approach, you’ll ensure a level of flexibility and extensibility while maintaining a prescriptive blueprint for exploring new capabilities. You may not know which LLMs you’ll need a year from now (you almost certainly don’t), but you can know how you’ll roll them out and implement them once you decide.
2023 showed us the “art of the possible,” and AI adoption will soar in 2024 as more organizations transform those possibilities into realities. As my co-founder Sinclair Schuller recently said: “This is not the time of system integrators; it’s the time of inventors and navigators.” That’ll be our north star in the coming year – will it be yours?