With tools like ChatGPT and DALL-E, AI tools are within reach for anyone with a computer and an internet connection to experiment with. But what do they really mean for our own professional and personal lives? More specifically, how can these tools be leveraged by any company now that they are so accessible?
With recent tech layoffs and focus on cost savings, we are often prioritizing immediate fires at the expense of getting to the next level of innovation. But the focus on building sustainable businesses with an eye toward profitability is a much welcome and needed shift. Generative AI tools are demonstrating possibilities in this space in a very simple-to-use format, pushing industries that have been collecting data (or not) to implement machine learning to further enhance their products and services.
Bridging the gap between research and application has also become easier and easier. R&D shouldn’t be walled off with engineers and product managers in their pods. The exciting realm of possibilities include the promise of closer collaboration between teams with each being able to highlight and use their domain knowledge and their specialized skills. If these easy to use AI tools can be leveraged to get to the next stage of innovation, that is the best kind of implementation. These tools provide scaffolding in a way that makes the jumping off point further out, and should alleviate fear of the unknown when it comes to AI. The use cases cut across industries and disciplines but businesses rightfully are grappling with what the next step should be for their individual cases.
What Engineers wished Business Leaders knew
At FourthBrain, after two years of training Machine Learning engineers and MLOps engineers with the mission of bringing more people into these careers, we asked ourselves and our graduates what is next. FourthBrain MLE and MLOps graduates have always been at the cutting edge of implementing Machine Learning projects in their jobs. They tried online courses and bootcamps before coming to FourthBrain for a structured way to learn along with their peers. They are the dream students, employees and leaders who embody lifelong learning,. While new roles and promotions with an average bump in salary of $27,000 were stellar achievements, what many craved most was interesting cutting edge work. Machine Learning engineers are who you want at the edge of innovation so when they tell us what they are looking for to take the field to the next level, we sit up and take notice.
When employers are vying for the scarce well trained Machine Learning engineers, making the work more interesting is their best bet at gaining an edge when hiring against FAANG (now MAANG?) companies. We conducted interviews at the end of 2022 with graduates as well as employers for what is next and a common theme that emerged was that engineers wished they were applying more Machine Learning at their work. They were ready, but wished leadership or executives were including more AI-influenced products in their roadmaps. Having an advocate at the Director or VP level championing the implementation of AI products makes all the difference. Bringing together product and data teams to design roadmaps and propose budget impacts is a win-win for the engineers, the company and their customers.
What does Investing in AI mean?
We also heard over and over again that Machine Learning was in fact an area that companies wanted to invest in. But what that meant in reality was nebulous. It is exciting that the potential of AI adding business value is recognized, and it is also understandable that the next step isn’t always clear. The good news is many companies and industries, especially those in financial services, healthcare, retail, telecommunications, real estate, oil and gas, etc. are primed and sitting on a lot of data ready to take the leap. Based on the Gartner Levels of AI Maturity1, companies at the first stage of Awareness versus companies already at the Experimental Stage may have varying action plans to get them to the next stage. Once AI starts demonstrating positive ROI for the company and investments start to pay off, the role of the executives in AI strategies becomes less of a driving factor. The plan or investment doesn’t necessarily mean adding more Machine Learning engineers at every stage but potentially repurposing teams and taking advantage of Quiet Hiring to stack people and resources against the most promising initiatives with the greatest ROI.
The Next Step
The good news is many companies are sitting on data without fully realizing the potential for what that data tells them about the next steps. For example, an edtech company knows a lot about their student behavior. When they study, for how long they study, where they get stuck or struggle, and even when they quit. The program design of an education product can take these insights to build out more support during the more difficult lessons, alert students to when an uphill struggle is coming and dedicate more time to it or even space out and build engagement for when someone will give up without being able to see the light.
When we interviewed companies in the space, the ones who were ahead and had a handle on their AI strategy had two differentiating factors: 1) at least one sponsoring executive who advocated for and brought the AI products to the forefront 2) deployment of ML models was demystified. The key to AI driven transformation is leadership2. Abandoned models were not a huge problem for the latter group. There was a distinct plan that allowed business units to have input into what types of ML models would bring business value. There were also Ops teams that helped with scoping so there was no waste with models that were stuck in the Jupyter notebook stage. They were vetted, improved and qualified with clear usability, feasibility and ROI before they got to the build stage.
An action plan that includes a to do list of how to make a roadmap happen. For some this may be an assessment of what level of AI maturity a company is at, or the plan to collect data if that didn’t exist already. AI will become more approachable, an executive doesn’t have to be a super coder to help align teams with the strategy. Culture and institutional knowledge will become more important and perhaps will lead to companies to look more internally to come up with and implement AI strategy rather than have an external consultant come in and do so. Building this capability will be the difference between those that level up versus those that are left behind. The AI revolution isn’t just coming anymore, it is here and getting out in front of it is key to not being left behind.