AI Ready Model – Labs

Artifical Intelligence (AI) is still a technique that is fairly young. This means that there is limited knowledge and experience available from how to actually implement and make us of the technology. AI has up until now mostly been implemented and tested with early adopters and tech savvy organisations. 

This means that there is a knowledge gap that needs to be addressed for your initiative to understand the technology, its requirements and how it might be implemented to enhance your business in various ways.

A good way to do this is to define an AI lab to try out its capabilities and requirements. In a lab environment everything is allowed and various techniques, models and infrastructure can be tested out. In this phase it is important to investigate which models are out there, what kind of pre-requisites there are for using the models and in what kind of infrastructure it exists. 

In the Labs phase you should also focus on what happens with the data you share and where it ends up in the end. You should also investigate how the model you intend to use have been trained and on which data. In some cases AI models have been trained on data that might not give you the answers you expect or desire. It is also desirable to investigate if there are solutions where you can inject your own data to train a model. 

Another challenge with AI models is that they sometimes don’t really know the answer to your question. In these situations they try to answer to the best of their knowledge, which might give you an answer that is not corresponding to the truth and just the AI ”best guesses”. All these things is something you need to take into consideration when implementing AI tech. 

That is why one of the purposes of running an AI lab, is simply to learn more and increase your organisations knowledge on AI and its pros and cons, to be able focus your initiatives resources on the right use cases.

Once you and your organisation starts learning more and gain valuable real-life experience from AI tech, you will also be able to start discussing how this can be utilised to support use cases in your business. Based on the learnings from the AI Lab phase, you can start defining use cases to try out. This is the second purpose of the AI Lab. 

To try out the technology to identify potential use cases. This is why another outcome/result of the AI Labs phase should be to identify a list of potential use cases that you can use for a ”Proof of Concept” (PoC) and the requirements to run them using AI tech. An AI Lab does not have to be dismantled after you have gone through this phase for the first time. On the contrary, the AI Lab should be a permanent place for members of your organisation to bring new ideas on how to utilise AI and make use of it in the organisation. 

With an AI Lab in place, your organisation will always know where to go to learn more about AI, its requirements and to try out new ideas.

At Redpill Linpro we have experience from running these AI Labs, both in our own organisation and for customers. Using AI Labs we have lifted both or own and our customers knowledge on AI tech and been able to implement a number of various use cases in different areas. 

With the use of our experience, we can assist in defining the structure for the Labs, mentoring, which tools and tech to use and procedures and methods required to move the initiative forward.

In the coming blog posts, we will dig a little deeper into how to create real value from your AI initiative and bring your organisation safely into the future.