Hybrid AI samples demonstrate its business value

Many organizations face two inherent AI-related problems: the need to automate at least some processes so that people can do innovative work, and the fact that dozens of existing chatbots are insufficient or error-prone.

Hybrid AI is not a new concept. The most common definition of hybrid AI is technology that combines symbolic AI (human intelligence) with non-symbolic AI (machine intelligence) to deliver better results.

Usama Fayyad, president of technology and strategic consultancy Open Insights, described machine learning as an iterative improvement of adaptive algorithms based on training data.

“Whether you’re building deep learning or next-level models [such as] probabilistic Bayesian models, you need a way to get the right training data, which usually comes from people who give the correct classification or interpretation,” Fayyad said. “Human interpretation and labeling are essential for learning systems ranging from machine-learned ranking in a web search engine to autonomous vehicle training.”

Concerning the fusion of deep learning methods with symbolic methods, Fayyad distinguished between procedural knowledge and declarative knowledge. Procedural knowledge means that people know how to do something without being able to explain it, while Explanatory knowledge can be pronounced. For example, many speech recognition and vision problems are procedural in nature, as they are difficult for humans to explain; therefore they are more prone to black box approachingor those that lack transparency.

“I see hybrid solutions being very important, both in dealing with procedural tasks and in addressing the current knowledge gaps,” Fayyad said. “In my view, the hybrid solutions are the right approach in almost all cases, especially if we want to explain and understand what the AI ​​does.”

The resources needed for effective hybrid AI

Successful hybrid AI examples demonstrate both domain knowledge and AI expertise to solving real problems† Without domain knowledge, the solution usually does not fit the problem. Without AI expertise, it can be difficult to understand challenges and understand what to do about them.

“End-users who are the intended consumers of some predictions can take an active role in a hybrid AI system as the ultimate decision makers about those predictions and can accept, invalidate or change any prediction based on their own personal and contextual knowledge.” said Fabio Pirovano, chief technology officer at Docebo, provider of AI-based learning suites. “However, to be effective, the AI ​​system must honor a ‘contract’ with the end user by making its predictions available to expert scrutiny within an acceptably rapid timeframe.”

In most cases, the effectiveness of hybrid AI depends on human judgment for training and optimization.

The technology support most needed is the ability to capture the final decisions of experts, either for offline analysis by the data scientists responsible for the original AI model or for use as additional training data that fundamentally improve the models

While many technology building blocks are available, building a coherent end-to-end solution is often a patchwork. Pirovano said he considers the most practical hybrid AI example today to be the human-in-the-loop kind, because technological tools needed for harnessing symbolic reasoning and statistical learning are relatively immature from a business standpoint.

Having the right mindset is also important, and that starts with identifying a business problem and then using the right technology to solve it – with or without hybrid AI.

“The key mindset is one where we have a deep understanding not only of the limitations of algorithms, but also of the deep reliance on data quality, availability and issues,” Fayyad said. “Most importantly, understanding whatever solution we come up with will require continuous feedback and rebuilding as the data, domain environment and requirements change.”

Common Benefits and Challenges

Today’s hybrid AI examples are most effective when humans and machines respectively do what they do best.

“Humans are good at making judgments, while machines are good at processing,” said Anand Masood, Chief AI Officer and Lead Architect at digital transformation firm UST Global. “The machine can process 5 million videos in 10 seconds, but I can’t. So let’s allow the machine [to] do their job, and if anyone smokes in those videos, I’ll judge how that smoking is portrayed.”

In most cases, the effectiveness of hybrid AI depends on human judgment for training and optimization. Otherwise a chatbot can degrade customer experience, for example. Therefore, the first major challenge is to staff hybrid AI projects with the right technical expertise. The second is removing both the lack of industry best practices about what hybrid AI systems should look like, and the lack of tools and frameworks to implement those best practices.

“The goal should be to understand when and how symbolic AI can best be applied and fruitfully combined with statistical learning models,” said Pirovano of Docebo.

Leave a Comment

Your email address will not be published. Required fields are marked *