The Commodification of AI: What it Means for Your Business, and How to Capitalise on It

Egor Kraev, Head of AI, Wise

Egor Kraev, Head of AI, Wise

To understand applied artificial intelligence (AI) these days, you have to think back to the time before the industrial revolution. Back then, if you wanted an iron plough, you had to ask the local blacksmith, an expert craftsman with years of training and experience. He would create a plough just for you.

But with the advent of the factories, the majority of the ploughs were made by machines or on assembly lines. Designing those machines and assembly lines required an even higher level of skill, but the factory workers needed much less expertise than the blacksmith to produce the plough. The level of skill needed to create the plough varied depending on where a worker was in the chain.

How does this Apply to AI?

Similarly, several years ago, creating a machine learning (ML) model required intricate craftsmanship. But now, more and more problems have out-of-the-box solutions that are, at the very least, ‘good enough’. For example, for making predictions from tabular data, gradient boosting approaches, such as XGBoost, LightGBM, and CatBoost, have clearly won.

Similarly, for working with natural language, there has been an explosion of models such as BeRT, now easily available through Hugging Face libraries.

Standard solutions for time series forecasting have lagged (with Facebook's Prophet being the exception), but lately, there has been a wave of big companies such as Uber, LinkedIn and Facebook open-sourcing their approaches. We’ve also seen a rise in ‘easy to use wrappers’ for academic work, such as PyTorch Forecasting.

Some domains are still not fully commoditised, such as natural language text generation (including chatbots) or image understanding and generation. Any packages that do exist here are not high profile, but there’s been huge progress already and it’s likely we’ll see these appear in a year or two.

Considering all of the above, it seems likely that in time, the majority of your data scientists will not be craftsmen, but assembly line workers.

It’s very possible that in three or four years, ML will turn from a job description to just another skillset, like SQL or Excel macros.

You still need some in-house experts to design that assembly line, but the rest should be able to cross-train from developers or analysts, or be hired from the large number of people attending ML courses. It’s very possible that in three or four years, ML will turn from a job description to just another skillset, like SQL or Excel macros.

Can You Rely Solely on ‘Out-of-the-Box’ Solutions Today?

In short, no. There’s still a lot of work to be done to advance the evolution of AI, and currently, not enough ML projects are making it into production. A significant cause of that is the lack of maturity in MLOps—which is the software stack needed to deploy, use, and monitor models in production.

Unlike the LAMP stack or the ELK stack in other domains, there is no single standard stack for that, and there are dozens of providers for specific components or component combinations of that stack, vying for supremacy. In short, there’s no set way to put these solutions into production, and that makes them hard to introduce, market, and implement.

The AI Infrastructure Alliance is trying to change that, but it's early days yet. Even the definitions of core terms, such as Feature Store, differ widely between providers.

How Can You Improve the Use of AI in Your Business?

First of all, you absolutely need a dedicated engineering team building your ML infrastructure, making it robust and at the same time allowing for smooth experimentation on real data.

The tension between robustness and agility is quite hard to manage. Therefore, the quality of these engineers, both in terms of building robust stacks and being able to think like a data scientist, is almost more important than the quality of the data scientists. Consider spending at least as much on your ML platform as on the data scientists themselves if you want the latter to have a business impact.

Secondly, free/open-source AI tooling is evolving at an incredible pace. Just like the Python ecosystem made MATLAB irrelevant for most AI work, non-free AI solutions will only thrive in certain specific niches. Examples are anomaly detection and compliance-oriented explainability— but the overall trend is towards free tools becoming good enough.

Finally, just like the blacksmith profession has not become extinct but rather reserved for special projects, there will always be AI problems that require expertise and custom solutions. However, it’s likely that they will increasingly be in the minority, and you will be able to have the most significant business impact by solving the standard ones. With ML tooling becoming ever more powerful and easier to use, the kind of data scientists you need to look for will also have either business domain expertise or fluency in the software development and MLOps domain.

We’re yet to see even a fraction of the possibilities that the commodification of AI will bring us. While there are challenges and complications as the technology evolves, there are also major opportunities for its application to grow.

Many of the major changes we see in the application of AI can and will happen in your organisation—if you do it right.