Will AI Become Like Organic Food?
Two years ago, everything was about mobility, and last year was all about the cloud. Those uber-trends are still with us, with a lot of runway left, but there’s something bigger in the room now – Artificial Intelligence. Of course, these trends are connected – especially in the collaboration space – so we have to pay attention to all of them. Each is relevant to collaboration in its own way, but it’s interesting to note how they’re getting increasingly abstract, and that doesn’t make UC-related buying decisions any easier.
Mobility is a broad concept, but we basically know what it means, and everyone has a wireless device that gives it tangible meaning. Cloud is ethereal by nature, but at least we know that our data and applications reside on servers in data centers. Where they actually are located is another question, but at least cloud is anchored to a physical object.
AI, however, is just a state of being, where we interact with machine-driven applications that automate things and hopefully make our life better. We may engage with AI via physical objects – a smart phone or a PC – but those are passive mediums through which AI is applied.
AI is now…a Thing
Just as UC vendors can’t get to the cloud fast enough, or become mobile-first fast enough, the same is happening now with AI. Every vendor in the collaboration space is building AI into their offerings, and pure play AI start-ups are the darlings of investors. There is a promising ecosystem coming together, and now that AI has found its way into the collaboration space, it’s become a thing, so to speak.
More so than mobility and the cloud, though, AI is very esoteric and remains the domain of data scientists. Our space is coming to AI late in the game, but in terms of a business opportunity it’s very early. AI has become well-established in other fields, and there’s a reason why it’s taken so long to arrive in the collaboration space. Workplace collaboration is more demanding than consumer applications, but also, voice-based forms of AI are complex, certainly in terms of what’s needed for collaboration. Simply put, our needs are challenging for AI to address, at least to the extent we understand it at present. That’s where things get a bit complicated.
I’ve noted elsewhere that speech recognition accuracy has now reached 95%, and that’s viewed as being good enough for most AI-driven applications. That’s an important milestone to make AI relevant for collaboration, and that’s certainly being reflected in the applications being developed for the contact center. To be fair, though, it’s worth noting that getting to 95% is the easy part. Closing the gap to approach 100% is much harder, and for AI to have true breakthrough benefits, it needs to get there. That may never happen, and we may end up settling for good or very good AI applications for collaboration, in which case, AI will be an exciting innovation, but not really a game-changer.
That aside, speech recognition is just one aspect of collaboration that AI has to be really good at, and it has a long way to go to become a core driver for UC. The technical challenges remain daunting, but the buzz is real, and most of us are probably rooting for AI. We have a general sense of the possibilities, and I think people genuinely want the benefits we’ve been hearing about to materialize. There’s little doubt that intelligent automation can make us more productive – workflows would be faster, knowledge workers would have more cycles to focus on complex tasks, error rates would decline, costs may come down, and we may even make better decisions.
You Need to Take an Intelligent Approach to AI
That’s my basic message here, as AI’s murky nature gives the sellers a leg up on the buyers. Unless you have a firm grounding in data science, it’s pretty hard to tell what’s really “AI.” You can’t see it or touch it, and given that most of us want to believe in the power of AI for good, we’re going to be naturally predisposed to any “powered by AI” messaging. Furthermore, don’t assume that the collaboration vendors you’ve been buying from all these years know much more than you do. Sure, they have lots of engineering horsepower and leading-edge AI partnerships, but until recently, AI hasn’t been in their DNA either. Just as you’re trying to figure out how to buy AI, they’re to figure out how to sell it.
I’m talking in broad strokes, but am trying to provide some fair counterbalance to the growing volume of messaging – and noise – coming from collaboration vendors, especially as they use AI for competitive differentiation. It’s a sound marketing strategy, but AI’s business value for collaboration is almost as difficult to pin down as, well, the business value of collaboration. For years I’ve been writing about this as a core challenge for UC, and until metrics emerge that can reasonably gauge workplace productivity, this adds another layer of complexity for decision-makers. If you can’t even define collaboration, how can you really assess the business value of AI for collaboration?
I have no doubt that AI will find its legs in our space, but currently, it’s hard to tell where AI is delivering something new as opposed to moving automation efforts faster along for projects that were almost there already simply based on software upgrades? In other words, when considering AI, ask the hard questions about where the real “intelligence” is coming from. While the messaging may scream AI, the reality may just be sexy window dressing on improvements made by a team of hard-working developers or software engineers.
Just Because it says “Organic”…
Doesn’t mean it’s automatically good for you or somehow healthier or more virtuous than what the masses are buying over in the produce section. You may feel better by making the choice to buy organic – and not think twice about paying the premium – but the vast majority of consumers have no idea how true any of this really is.
The cynics out there will agree with my view that whatever positive affinity you have towards organic food, this is first and foremost a marketing tactic that food producers understand very well. When standards are loosely defined, as with anything else, it’s buyer beware, and once the “organic” label is placed on the product, the grocer has a free pass to charge more because of the higher perceived value.
It’s no different with AI, and while I don’t think vendors are out to fool the buyers, they know the “AI” label has marketing power, and they’re not afraid to use it. In these early days where everyone is gunning for first-mover advantage with AI for collaboration, you just have to ask enough good questions to make sure you’re getting enough real AI for your money.
Furthermore, as the market matures, there’s a risk that AI will become so over-used that, like “organic,” it will lose meaning and its perceived value will diminish. So, when you see AI being applied to everyday tasks like making phone calls, you’ll know it’s gone too far, and will need to be even more vigilant determining where the real business value is. On the bright side, of course, AI may well evolve to elevate these mundane activities to experiences we can’t even imagine – and that would be really cool. That’s the AI we’d all be happy paying a premium for, but I think we’ll be waiting on that for a while. Until then, this might be a good time to read Arthur C. Clarke and watch 2001: A Space Odyssey with fresh eyes.