Accenture just released its Technology Vision 2024 report that suggests AI will unleash the next level of human potential. It specifically calls out the rapid pace of innovation in generative AI, agentic computing, spatial computing, and new human interfaces.
The big deal is that the combination of these new capabilities will make computer interfaces more human rather than requiring users to become more machine-like. In the long term, this could upend the technology landscape as much as the mouse, cloud, and mobile waves before it. In the short term, enterprises should focus on modernizing their data systems and improving their human talent to take advantage of the new opportunities.
Here is a bit more detail about the most important trends:
A match made in AI: re-sharing our relationship with knowledge
Generative AI technologies will play an essential role in re-organizing knowledge in ways that facilitate human-like reasoning rather than combing through a mass search engine result. It is essentially transforming the search experience into a synthesizing experience for business users and consumers. Accenture predicts that by 2029, an AI advisor will receive more search traffic than traditional search engines.
However, these large language model advisors will require a data foundation that is more accessible and contextual than ever. Accenture says:
The knowledge graph is one of the most important technologies here. It’s a graph structured data model including entities and the relationships between them, which encodes greater context and meaning. Not only can a knowledge graph aggregate information from more sources and support better personalization, but it can also enhance data access through semantic search.
The report cites how Cisco’s sales teams turned to a metadata knowledge graph using a Neo4J database to halve salespeople’s time searching for data, saving over four million hours annually. In this case, they saw these gains without the help of generative AI. Data mesh and data fabrics should also be investigated outside of knowledge graphs as they update their overall architecture. Large Language Models (LLMs) will also play a big role in fleshing out ontologies that find connections between entities and their relationship to each other and populate the graph database.
Meet my agent: ecosystems for AI
The first generation of LLMs is already helping to build individual AI agents across enterprises like Bloomberg and Morningstar. The next wave of innovation lies in building out a trusted ecosystem of domain-specific agents. Accenture reports that 96% of executives agree that leveraging AI agent ecosystems will be a significant opportunity in the next three years. Early examples include AutoGPT, BabyAGI, MetaGPT, Google’s LATM, and the ChatGPT plug-in ecosystem.
However, Accenture cautions:
But there’s a catch: there’s a lot of work to do before AI agents can truly act on our behalf or as our proxy. And still more work before they can act in concert with each other. The fact is, agents are still getting stuck, mis-using tools, and generating inaccurate responses – and these are errors that can compound in a hurry. Without the appropriate checks and balances, agents could wreak havoc on your business.
In the short term, it’s important to build the scaffolding agents need to gradually earn an organization’s trust. This includes weaving together LLMs, human agents, and existing enterprise apps while keeping an eye on governance and security. Human experts will also need to embed rules, knowledge, and reasoning skills, test them, and decide when and where they can be trusted on their own. It’s not just about teaching agents to master new skills. We need to ensure they build a world we want to live in.
The space we need: creating value in new realities
The industry has made significant progress on tools, processes, and standards for extending user interfaces from 2D screens to 3D environments. This includes spatial computing, metaverse, digital twins, augmented reality, and virtual reality. These will help fuse digital and physical worlds to create new experiences for business users, consumers, physical product developers, front-line workers, and engineers. Spatial apps will help convey large volumes of complex information, give users agency over their experience and allow us to augment physical spaces.
A new computing medium is exceptionally rare, and so a tipping point lies ahead. Spatial computing could grow to be as groundbreaking as desktop and mobile, ushering in a new era of technology innovation. But to succeed, enterprises need to rethink their position on it, starting today. They need to get out of the slump and recognize this moment for what it is. The tools are more ready every day – how you apply them is what matters now.
Apple’s entry into the market signals that a new technology medium has arrived. Qualcomm’s Snapdragon Spaces and SR SDK will facilitate more economical variants from smaller competitors. Work on new 3D standards like the Universal Scene Description (USD) and glTF will play a significant role in connecting workflows and experiences across various apps, services, and workflows.
Our bodies electronic: a new human interface
Innovators have been making rapid advances in new human interfaces around AI-powered wearables, brain-sensing neurotech, and eye and movement tracking. These could unlock a better understanding of us, our lives, and our intentions to enhance the way we work and live. An Accenture survey found that 96% of executives agree that human interface technologies will let us better understand behaviors and intentions, transforming human-machine interaction.
However, this brings a variety of new privacy and security challenges that enterprises need to consider. Accenture suggests:
Think of it like mobile device management for humans: we already know how to control what mobile device data stays local or gets sent to the cloud, but the stakes are higher and more complex for sharing humans’ biological, behavioral, and sensory data. While rule-based approaches can lay the foundation for data-sharing systems, humans will need more flexible and interpretable safeguards to maintain control of their own data. The data will also need to be interpreted for them, so that there is no ambiguity about what access they are granting.
Apple’s new visionOS may promise a seminal advance in a new eye-tracking user interface. Accenture is also bullish that brain-computer interfaces (BCI) are just around the corner with significant advances in low-cost hardware and algorithms. This could extend the use of the technology beyond scientific research and geeky meditative assistants to practical enterprise use cases like improving training or speeding processes. One interesting use case was an app that could monitor a person's brain while viewing airport security scans to flag firearms at a rate of three images per second with high accuracy.
A careful reading suggests that many of these innovations are still a few years away from prime-time and safe adoption. Each new technology also comes with a variety of new security, privacy, and regulatory risks that need to be considered in building trusted, secure, and safe systems that deliver real business value.
But when they do finally land, enterprises that master them will achieve a significant advantage in their industries. All four areas are definitely worthy of some investment and experimentation to see how they might fit into the enterprise when the ecosystems around them start to mature.
But don’t dive head in just yet. Enterprises will need to focus on improving their human talent and integrating their data infrastructures together to see the most significant benefits of this coming wave. On the talent side, all four innovations will require substantial investments in the humans tasked with building trust and applying them to specific business cases.
On the data side, this is not just about wiring data systems. They also need to weave data's context, significance, and meaning across different use cases using knowledge graphs built across graph databases, ontologies, and taxonomies. Generative AI could play a significant role in the short term in automating the process of building out these knowledge graphs.