To help frame No Jitter’s year-end 2023 retrospective coverage, No Jitter (NJ) created a word cloud from the weekly No Jitter Roll (NJR) articles – as shown in the following graphic. As the word frequency list shows, “AI” was used twice as often as “customer” – which certainly emphasizes where vendors in the enterprise communications, collaboration and contact center space were focused this year.
Created with: Free Word Cloud Generator
When Generative AI (gen AI) first debuted in 2023, most of the focus was on how naturally they were able to understand and generate text replies, images, music, etc. As 2023 marched on, gen AI use cases extended into streamlining the creation of rote content and/or removing some of the “heavy lifting” from laborious knowledge-based processes, such as writing email, summarizing transcripts, assisting with designing a conversational dialog flow, etc. Many of No Jitter’s thought leaders and contributors provided their varying perspectives on the multitudinous uses of gen AI, along with recommendations on how to implement, as well as the hype, reality and potentiality, of the emerging technology.
“Too many conversations in boardrooms are around ‘how can we make money of generative AI?’ – this is just ‘fear of missing out,’” said Natalia Modjeska, Research Director, AI and Intelligent Automation with Omdia. “Instead, they should be focusing on business challenges and whether AI, and in this case, generative AI, could help solve those. Gen AI is just one of the tools in the enterprise toolkit, and it’s not always the right or the only solution.”
In a recent NJ article, president and principal analyst at Metrigy, Irwin Lazar echoed Modjeska’s point. “As companies get their hands on generative AI assistants they are beginning to focus on ROI, especially when evaluating vendors who charge for AI assistants,” Lazar wrote.
What did a year of excitement and experimentation produce? See below, for some of the biggest trends in gen AI this year.
NJR has covered many vendors who have all launched some type of generative AI solution in 2023: GoTo, Five9, Cognigy, Zoom, Genesys, Google Cloud, AWS’s Amazon Connect, Webex, Observe.AI, Cyara, CallMiner, Twilio, Salesforce, Emplifi, Luminoso, Minerva, ASAPP, ServiceNow, Dialpad, Kore.ai, Omilia, Cresta, and, of course, Microsoft who, relative to all these other vendors – as shown in the Word List above – dominated the headlines with its space-defining Copilot which uses OpenAI’s large language models (LLMs).
“Literally, every CX provider is working on the technology, so it would take a lot of space to name them all. As I work with enterprise IT and CX leaders, they aren't as concerned with the underlying technology or even how it works. AI- and CX-related innovations are moving so quickly, they can't keep up,” said Robin Gareiss, CEO and Principal Analyst at Metrigy.
In the contact center, generative AI-based capabilities were introduced that were built on that – e.g., summaries that assist contact center agents during and after customer interactions. Contact center managers also benefit from AI-generated summaries of calls, trends, etc. And all those summaries – in this example – cut precious seconds and minutes out of those pre-, during and/or post-interaction activities.
Enterprise IT and CX leaders “have gotten to the point where they say to their technology partner: ‘Here's my problem or opportunity. Tell me how to address it,’” said Gareiss. “So, the relationships with their technology partners will become more vital – and more sticky – than ever.”
Check out the following for NJ’s coverage of this trend throughout 2023:
In the unified communications and collaboration (UC&C) space many companies also integrated generative AI. For example, Microsoft 365 Copilot, Zoom AI Companion and Webex AI Assistant, all provide “assistance” – summaries of chats, email and meetings, draft composition, idea generation, etc. Many companies have made it clear that these capabilities are just the beginning.
And, of course, Microsoft’s 365 Copilot is integrated across its portfolio of products and in Windows 11 itself. According to “What to Expect from Teams Now,” by Kevin Kieller, Co-Founder and Lead Strategist with enableUC, “Microsoft provides a complete set of tools to create, deploy, and extend AI applications and models. Microsoft uniquely has a complete set of data consolidation and management tools, the compute infrastructure to train and deploy sophisticated AI models, and the developer tools to create new models, extend existing models, and embed AI in existing applications.”
Zoom’s approach appears to involve an effort to redefine collaboration, as evidenced by the introduction of Zoom Docs, while also providing integration into Microsoft’s productivity suite. Webex provides those integration, too, as do most of the other UC&C players – from RingCentral to NEC.
Check out the following for NJ’s coverage of this trend throughout 2023.
Many companies standardized on a single LLM provider, mostly OpenAI which entered 2023 with momentum and a key investor/customer/partner in Microsoft – which runs the Azure OpenAI Service. To get access to the ChatGPT family of models, enterprise and/or vendors must work with Microsoft or OpenAI. (Note, too, that Cohere’s enterprise AI model, Command, will be available through the Azure AI Model Catalog and Marketplace.)
There are multiple LLMs available from various companies such as Anthropic, Meta and Google Cloud, to name a few. Currently, one challenge enterprises face is choosing the best LLM for their use cases.
AWS took a different approach and launched Amazon Bedrock, which is a managed service that provides customers with access to multiple models, as well as a new service that helps customers trial the different models available through Amazon Bedrock.
Webex and Zoom, among others (as cited below), have taken a federated approach to LLMs – which basically means they abstract the choice of which model is used from the customer.Zoom’s CTO, Xuedong Huang, recently wrote that Zoom’s federated approach allows them to adhere “to a cost-effective strategy that first employs a lower-cost LLM most suitable for the task."
Here are more examples of how various companies have approached of enabling access to LLMs in 2023:
