Embracing the Future: Predictions for AI in 2025
Happy New Year! As we step into 2025, the world of artificial intelligence is changing faster than ever. Generative AI, though still new, has the potential to completely transform how we work. Here at Synozur, we’re watching the trends that will shape AI in the coming year.
Last year, Katya Linossi, CEO of ClearPeople, asked me to share my insights as part of their eBook The Trends Shaping Knowledge in 2025. (ClearPeople builds enterprise solutions for Microsoft 365 to unlock enterprise knowledge and drive business growth.) It was an honor to be published alongside other industry luminaries. My predictions include:
"Omni AI" that works across all platforms
Anxiety about powerful AI agents
The rise and rise of personal knowledge chatbots
Let’s look at them in more detail:
Omni AI Integration
Every software vendor is adding or enhancing their AI chat interfaces. Each can generate answers from their unique knowledge store. That’s leading to AI proliferation. If you need to prepare for a client meeting by going into Copilot, Salesforce, ServiceNow and SAP, that’s a lot of context switching. Haven’t we learned past lessons about the cost of task switching? Synozur research, based in part on data from Harvard Business Review[1], shows that the productivity drain from knowledge worker task switching will reach a staggering $279.5 billion this year.
There’s an emerging path to improve both AI quality and ROI. Businesses will increasingly demand AI that develops answers on data integrated from multiple sources and disconnected AI chatbots, far beyond single-source systems like Gemini and Copilot. Let’s call it “Omni AI.”
Omni AI is based on two approaches which are not mutually exclusive:
Content integration. GenAI chatbots need not be limited to a single source if information. Most organizations have dozens of sources of data – a typical $1 billion organization may have more than 50[2]. Customers need a manageable way to create an indexed source of data for LLMs and RAG (Retrieval-Augmented Generation). Microsoft has made some initial steps with its Graph Connectors. I predict an emerging ISV will successfully introduce a next generation solution to unify, enrich and secure cross-platform information for GenAI.
Federated chat interfaces. From a user perspective, it doesn’t matter whether their chatbot has a unified “library,” or if it asks a team of other local AI agents to work on their parts of a problem before combining them into a cohesive whole.
Omni AI can create cohesive knowledge systems, empowering better decisions and employee effectiveness. Organizations that adopt Omni AI solutions will experience greater productivity gains than those still using 2023-era AI.
Anxiety and Skills-Based AI Autonomous Agents
In 2025, you can expect AI-powered autonomous agents to take digital automation to the next level. These agents will go beyond just managing workflows and robotic process automation (RPA). They’ll handle complex tasks like managing incoming emails to onboard new customers, handling contact management, generating contracts, and scheduling meetings.
Recently, EY's AI Anxiety in Business Survey revealed that 65% of employees are anxious about AI replacing their job[3]. The fear of AI job disruption will remain much larger than the disruption itself – it’s much harder to automate complex human interactions than we think. But managing that fear will be essential for organizations looking to move quickly to embrace AI solutions.
Personal Knowledge Chatbots
In many organizations, our knowledge maps are personal. We try to ask the “best” expert based on what we know they know. For example, Sloane knows the most about the Pittsburgh warehouse design since she’s the manager, so everyone asks her for details.
Personal AI for consumers and business are converging. For example, Character.AI can impersonate anyone – even John F. Kennedy - with a small voice sample, but it doesn’t know or retain anything unique about people. It initially told me I was born in San Jose, CA (Rockville Centre, NY); that I attended University of Georgia and UC Berkeley (see below) and that had three children – Connor, Keegan & Ellie. Not even close.
In 2025, business AI like Microsoft Copilot will it easy to configure an agent that’s grounded on your unique digital workplace footprint (email, chats, documents, meetings etc.), blended with external sources like LinkedIn or even Ancestry.com. In theory, it will “know” what you know. I tested this by building a Copilot agent grounded on my business and personal OneDrive content (which includes my bio and resume.) Copilot got my education correct – Boston College (Go Eagles!):
Copilot after learning about me…
So, Sloane’s chatbot agent may soon be able to answer frequent questions on her behalf.
According to AIM Research[4], the global chatbot market is worth $7 billion in 2024, and growing at 24.32% annually. However, personalized knowledge chatbots will face concerns about ethics and quality. Sloane’s chatbot may answer instantly, but is it correct? There’s a big gap between organic and digital knowledge, and there’s no replacement for human judgment, intuition, and creativity. These agents will remain a curiosity in 2025.
Wrapping Up
Looking ahead to 2025, it’s clear that AI will continue to change the future of work. At Synozur, we’re here to help you navigate these changes. By understanding and leveraging the latest AI trends, you can set yourself up for success in the digital age.
Join me on February 11 at 8:00am for our free webinar How to Succeed with your Copilot Rollout. Based on dozens of early adoption customer programs at Microsoft, we’ll cover essential strategies and practical tips to ensure a successful Copilot implementation, from preparation and content readiness to pilot programs and full-scale rollout.
Join us on this journey towards the future of work, where AI isn’t just a vision, but a reality we’re building together. Learn more about our strategic advisory services and how we can help you achieve digital excellence at www.synozur.com. Thanks.
References
[1] https://hbr.org/2022/08/how-much-time-and-energy-do-we-waste-toggling-between-applications
[2] Delivering Business-Critical Solutions with SharePoint and Line-of-Business Data, Microsoft, 2013.
[3] https://www.ey.com/en_us/newsroom/2023/12/ey-research-shows-most-us-employees-feel-ai-anxiety
[4] The Chatbot Market: Trends, Growth, and Key Players (AIM Research, 2024)
