By Ann Maya
The last major technological revolution — what we might call the “smartphone revolution” — put a small computer in the hands of virtually everyone. It completely democratized our ability to connect, collect, and communicate information to shape our world through digital means. It nearly demolished the barriers of geography and access to knowledge, helping lead us — with contributions from the innovation of cloud storage and massive leaps in processing power — to where we are today: the AI revolution.
In this blog, we’ll take a look at some of the broader trends triggered by the rise of AI and the insights revealed in a recent BARC global survey, sponsored by Boomi, “Optimizing Your Architecture for AI Innovation.”
Computers That Teach Themselves
Advances in AI have spawned an innovation cycle that is distinguished by something very cool — teaching the computer how to teach itself. I know what you’re thinking. This phenomenon has been happening for a long time. And you’re right! Google translate? That’s AI. Siri? More AI. Your bank’s chatbot? Yep, you guessed it. That’s AI, too. But what really threw accelerant on this bonfire of intelligence was the transformer.
This changed the game by allowing AI models to take statistical best guesses through the use of neural networks and applied layers of probability to “pay attention” to mass amounts of data in order to generate a response much like a human does. The nuance is the ability to apply contextual relevance to massive amounts of data (far beyond human capacity) all at once and produce an answer to a query that is “good enough” rather than search for a perfect response.
Good Enough Becomes Pretty Darn Good
The transformer is what enables the model to quickly find context to generate content, and by taking in new data and adjusting context as it goes, these models can learn. Genius! And learn they did. In fact, “good enough” has become not just pretty darn good but original. Add some fine-tuning, and it’s been able to create very human-like code, art, writing, and much more. From that moment, the world changed. There’s no doubt about it. You either accept AI, embrace it, understand it, all of the above, or get left behind. AI set technology on fire. We should not underestimate the speed at which change will happen.
Keep Calm and Carry On
But, before going all in on AI, it might be best to assess where we actually are in realizing business value from AI. As the BARC report points out, “This is an early-stage market — don’t let haste and FOMO (fear of missing out) drive your strategy.”
Today for many, going “all in” may mean first ensuring that your systems and data are in order. One of the key takeaways from the BARC report emphasizes this: “AI has surfaced a renewed requirement for highly performant data systems.” So, completing your digital transformation initiatives, including an architectural review, are recommended to prepare your organization to take full advantage of the innovation AI offers. Also, for virtually every organization entering the AI innovation whirlwind, there will be a “skills gap.” BARC recommends addressing this gap early in your AI journey.
Preparing Your Business for AI
There’s a lot to consider when successfully implementing AI solutions in a business. For modern solutions, you need modern, proven technology because efficiency and quality are imperatives for speed, and speed is crucial to winning in the business market place. For tips on preparing your organization for AI, check out Boomi’s Six Considerations To Help Speed Your AI-Readiness.
At Boomi, we’ve helped organizations thrive in digital business for decades by providing market-leading intelligent automation and integration solutions that span application, device, events, system and data integration, master data management, B2B/EDI modernization, and process automation. We connect the data that provides the business context for your AI systems. The Boomi platform is helping many organizations get AI-ready.
We’ve also developed our own AI engine. From the initial release of our cloud-native platform, we started collecting de-identified metadata that shows how our customers connect their systems. Over the years, we’ve seen more than 200 million integrations, which amounts to about 70-plus terabytes of data! This knowledge fuels our Boomi AI solution, which adds value across our entire unified platform.
Many key factors that impact AI aptitude and readiness were top of mind for participants in the BARC survey, which the report explores in depth.
Read “Optimizing Your Architecture for AI Innovation” for insights that will help inform your plans and discussions around the best path to practical AI.