The world of data science has become so complex, especially given the rise of AI, that no individual can be an expert in every aspect of the field. As a result, both individuals and companies must constantly seek collaboration with others to get work done. However, there are different types of collaborative partners, and each type can play a key role if used properly. We’ll dig into two primary types of collaborative partner and then make the case for which type of partner you want to be given the trends in AI today.
Task-Based Collaboration
Task-based collaboration is a very tactical and transactional approach. This model is about seeking someone to carry out specific tasks in the immediate term without any guarantee of future collaboration. In day-to-day life this would be the equivalent of hiring a neighborhood teenager to mow your lawn once or twice while you’re on vacation. You need your lawn to be mowed, and the teen is happy to make some cash. However, neither of you expect or care about a deeper relationship at the time the mowing is agreed to. Of course, if the teen performs well, you’ll be more likely to hire them again for your next vacation.
In the world of data science, task-based collaboration can take many forms. It might be hiring a contractor to configure a new computing environment for you or hiring someone to code a component of an analytical process that has already been defined. It is about having someone fill an immediate, well-defined tactical need. You need confidence that the partner has the necessary skills, but as long as they do, there is little risk involved to either party.
Relationship-Based Collaboration
Relationship-based collaboration is a more strategic partnership where both parties have more skin in the game. This model is about finding a partner who owns and manages part of your work with a mandate to self-manage for success. In day-to-day life this would be the equivalent of an experienced full-service landscaper. Rather than just tactically mowing your lawn a few times, they own the ongoing mowing of that lawn as needed. In addition, they own other aspects of lawn care such as weeding and fertilizing, in addition to maintaining all landscaping. In this case, you may not even know all the tasks required to maintain your various shrubs and flowers. Instead, you trust your landscaping partner to get it done.
In the world of data science, relationship-based collaboration can also take many forms. It might be hiring a consulting firm to manage your systems, or to own the creation and maintenance of a range of models, or to help manage a variety of major cross-functional projects. The key is that the partner has a much bigger influence on the outcomes because they are performing higher value, more strategic tasks. Additionally, the partner is being trusted to independently manage aspects of the work required that you as the person or organization hiring them may not understand. In other words, you don’t have perfect specifications for what the partner needs to do because you aren’t even sure what to specify! This type of relationship requires a high level of trust.
Comparing The Models
Both the task-based and relationship-based models have a place when seeking collaborators for your business. Sometimes you just need some helping hands to get things done. It isn’t that you and your team couldn’t do it, but you would rather focus on other things. Other times, you need substantive help from someone who can take ownership and who offers skills that you don’t have. Most organizations use a mix of both types of collaborative partner. If you’re looking to be a partner to others, where should you focus?
I have personally always defaulted to wanting to be as involved as possible. Thus, I’d rather focus on the strategic items like the full landscaping contract rather than the tactical items like mowing a lawn a few times. That’s just my preference, however, and plenty of people make plenty of money with tactical, task-based relationships.
However, with generative AI quickly rising, the tactical activities are much more exposed to either being replaced completely or shrunk in scope as AI handles more of the tactical items. If all you offer is basic coding, that’s easy to replace. If you offer to scope, design, and document what needs coded, that’s harder to replace.
Conclusion
There are merits to both types of collaboration, and you and your organization likely have need for people who offer both. However, over time more and more of the tactical, task-based work will be automated through AI. That’s not a problem as the owner of the work. However, if you’re the one looking to be a partner to do the work, it’s a big problem if much of what you offer can be automated and replaced with AI.
Thus, my recommendation is that you focus on building the skills required to be a relationship-based partner regardless of whether you plan to have your own business or plan to collaborate within your organization. The demand for strategic, relationship-based collaboration skills is going to remain steadier than the demand for task-based skills. Over time, AI may eat into relationship-based tasks as well, but there’s a lot more time left because higher level thinking and problem solving skills will be much more resistant to AI replacement.
Originally posted in the Analytics Matters newsletter on LinkedIn
The post Data Science Collaboration In The Age Of AI appeared first on Datafloq.
