As 2022 comes to a close, it’s a great time to reflect on the themes that emerged over the past year, both in the market at large and with our customers and community specifically.
Our survey of CDOs, released in February of this year, forecasted the importance of AI skills and the choice of companies to build them in-house in order to be able to deliver business impact with AI. Our most recent research, announced this week, found that companies are increasingly investing in AI and NLP to be able to reduce costs, drive growth and offer a competitive advantage.
While we’ve covered a lot of ground on the blog this year, we see those same topics reflected in the posts that resonated most with our readers: overcoming skill gaps, understanding the differences between different AI approaches, and finally, the need for AI that is understandable and explainable.
We’ll be sharing more about this new research in a future post, but in the meantime, here are our most-read and shared posts from 2022:
Rescuing Machine Learning with Symbolic AI for Language Understanding
Machine learning is great for pattern recognition and programming basic language understanding tasks. However, when it comes to understanding the nuances of language, especially for your unstructured data, it needs a lot of help. We look at what it really takes for machines to learn and be trained and what you need to know when evaluating #ML for your real-world, language-intensive use cases.
Overcoming AI Skill Gaps with Accessible Technology
When the supply of talent with specific AI and ML skills falls well short of demand, how can companies address their skills gap? The solution may be less about skill development and more about your choice of AI technology. Low-code and no-code platforms are an opportunity for less skilled employees to be exposed to complex AI technology. Discover how an accessible AI approach reduces the barrier to entry and makes it easy to train AI models right out of the box.
3 Best Practices for Implementing Hybrid AI to NLU Applications
In AI, there’s more than one way to train data. Hybrid AI, also known as composite AI, merges two traditionally opposed approaches to natural language processing: machine learning and symbolic AI. See how they work together to accelerate the enormous training data requirements of machine learning methods, and how to find the right mix to manage the complexity of your business problem.
Implicit vs. Explicit Knowledge for Language Understanding
The results you can expect from use cases like text analytics and smart chatbots depend on your approach to understanding language. This look at how different approaches tackle implicit and explicit knowledge shows why you need to know exactly how AI systems learn and acquire knowledge, and demonstrates the advantages of an approach that uses both symbolic AI and machine learning.
From Black Box to Green Glass: The Responsible AI Imperative
If you’re evaluating AI, you need to know how it can produce the useful results that you seek. But what does responsible AI look like and how can you achieve it? CEO Walt Mayo looks at the challenges of large language models (LLM) for responsible AI, why we don’t have to sacrifice responsibility for quality, and how we can move away from a black box approach.