👩‍💻 Data Strategy, Ethical Data Collection, Hiring Soft skills vs. Hard skills

This week, we spoke to Vince Walsh, Director of Data Science & Strategy at Starday Foods on: 

• The role of a Data Strategist in facilitating data scientists' work, ensuring data availability and organization. 

• Customer Data Ethics: Using public data ethically; prioritizing anonymity and avoiding personal identification. 

• Skills to prioritize when planning and hiring for a data science team. 

Q&A below.

💼 AI in Business 

AI Will Eventually Fade From View, By Design 

Forbes explores how commoditization in the IT realm has been a longstanding trend, where products and technologies become widely interchangeable. 

• Artificial intelligence (AI), particularly generative AI, is increasingly becoming ubiquitous and available at low or no cost, raising questions about its potential commoditization. 

• Nicolas Carr's concept of IT as a utility, explored in "Does IT Matter?" suggests that widespread access to technology levels the playing field, diminishing competitive advantages. 

• While some argue that AI will follow suit and become a standard tool rather than a differentiator, others point to the success stories of tech-savvy companies like Amazon and Uber. 

• The debate centers on whether innovation and forward-thinking application of AI will continue to provide competitive edges. 

• Current trends show organizations mainly focusing on tactical benefits of AI such as efficiency, productivity, and cost reduction, rather than strategic innovation and growth. 

• AI's role may evolve into an invisible yet transformative force in businesses, akin to oxygen, influencing daily operations and decision-making. 

• Despite predictions of AI fading into the background or being commoditized, there are opposing views suggesting that AI's integration into daily life may become more pronounced. 

• Presently, discussing AI implementations can provide businesses with a competitive advantage, showcasing technological prowess and commitment to enhancing services and products. 

💬 Q&A - Data Strategy, Ethical Data Collection 

My first question is related to your work in data strategy. How does a data strategist work with a team of data scientists? What does that look like? 
 
It's about enablement and making sure data scientists have what they need to get their job done. 

This field is still new to a lot of people. There will be instances where data scientists are hired, and maybe there's not the data they were promised or it's not ready to have models built on it. It becomes hard for data scientists to do their job. 

The role of the data strategist is really to make sure that all the other pieces are in place so that the data scientists can get things done and they're not spending all their time acquiring data, organizing data, doing the job of either a software developer or data engineer. 

But all that stuff is getting done before they get involved and ideally even a strategist builds out an idea of what can be done with the data and then the data scientist comes in and does the testing and figures out if that strategy is viable and builds out in the actual solution that can be you know deployed in an automated fashion. 

I've heard a lot from the data professionals that infrastructure is a big challenge to get right. How does the role of a data strategist fit in in that case? 

You know, it can be. 

I would argue, if you have hired a data strategist and don't really have your infrastructure built out you have more problems than you think you do, or than you know you do. 

Having a data platform in place is extremely important to having a successful data science program. 

If you skipped all that and expect a data strategist to come in and fix it, you have to be looking for a specific type of data strategist who really is more in tune with data engineering than they are with data science (they're more at the beginning of the spectrum than at the end). 

Some people can do both. And in that case, you'd need to be an early-stage company where there just hasn't been anything set up. But if you're like an established company, you probably aren't going to be able to put that on one person to come in and clean up what has been done previously. 
 
How should businesses balance the need for customer privacy when collecting data? 

When you post something, occasionally either your handle or your picture is also attached to that. We don’t use any of that information for targeting or furthering our business. We are more interested in just the content of the data. 

Again, in the public sphere, it is about not creating a database on a particular person, or entering information in that's attached to that person and then reselling that information like a lot of companies that have access to your information actually do, you know, like credit card companies, Facebook, etc. 

I believe it's ethical to use the data that people have put out into the public, but not create entries on who did what and who said what and then sell that information out there to other people. 

So keeping it anonymous basically.  

Exactly. we don't pull in personally identifiable information. So it makes it a little easier to deal with.   

I was reading through a few case studies on the Starday website and I saw mention of ‘zero party data’. Is that something that the customers are incentivized to share? How does it work? 

It's a term for data that was handed to us with no real agreement either way. it's not really useful for us in terms of machine learning models that we have because it such a small data set, but it is useful for identifying what people are using our products for and answering any product development questions that we may have (from feedback forms). 

So if someone sends us a picture or if they give us information through a survey we can understand if our packaging is expressing the things that we wanted to express, we get a sense of how people are actually using our products. We then know how to market better. 

When you're hiring someone for your data team, how do you weigh their soft versus hard skills. Which do you place more importance on? 

That's a really good question and I think when I look at hiring, I am really focused on what I want from that role. 

When we're hiring someone, before we even make a job description, we build out a 90 day plan and we break it down into 30, 60 and 90 of what we expect this person to accomplish. It's a good exercise for the hiring manager too, as it helps us identify the type of person we need. That helps us build the job description and then with those two things in place, we can think about -  

What is more important in this role? Is this something where we can teach them the technical parts?  

I think there's like a baseline of technical skills that I need to see right. I don't want anybody coming in, no matter how great of soft skills that they have if they don't already know SQL, right? That is the base for a lot of products that are out there. 

Do I need every person to have experience with our entire data stack? Absolutely not. Because there're so many tools out there, I don't expect to find a person who has the exact technical skills that we're looking for. 

So, when I'm doing my hiring for the data team, I probably have a higher threshold for technical skill than most people, but it's not the only thing that's important. I am looking for at least some soft skills in most roles, but it always goes back to the 30-60-90 plan and the JD for that role. 

🌐 From the Web 

Saudi Arabia Spends Big to Become an A.I. Superpower 
Saudi Arabia invests heavily in tech, hosting events like Leap conference to lure international firms. 

Ethical A.I. and Innovation Are 'Two Sides of the Same Coin' 
Tech CEOs at the 2024 TIME100 Summit emphasize ethical AI innovation, stressing the symbiotic relationship between regulation and advancement for a better future. 

Big Tech keeps spending billions on AI. There’s no end in sight. 

Big Tech's AI spending skyrockets, with billions more planned, straining chip supply and U.S. power grid. Google, Microsoft, and Meta lead the charge, reshaping the tech landscape. 

🏳️Ethical AI 

The ethics of advanced AI assistants 

A paper from Google DeepMind explores the future that may see widespread integration of advanced AI assistants, revolutionizing daily interactions and societal norms. 

• General-purpose foundation models are enabling AI assistants to perform various tasks, potentially serving as creative partners, tutors, planners, and more. 

• Ethical and societal considerations surrounding AI assistants, including value alignment, safety, misuse, and impact on economy and environment, are crucial. 

• Research indicates profound impacts on users and society, influencing work, education, creativity, and social interaction. 

• Human alignment with AI assistants is vital to mitigate risks of accidents, misalignment with user values, and misuse. 

• Communication in natural language raises concerns about trust, privacy, and appropriate relationships with AI. 

• Cooperation among AI assistants is essential to avoid collective action problems and ensure inclusivity. 

• Comprehensive evaluations are needed to anticipate risks and inform responsible decision-making regarding AI assistants' deployment. 

• Current decisions by researchers, developers, policymakers, and the public will shape the future of AI assistants, emphasizing the need for coordination and cooperation for beneficial outcomes. 

🤖 Prompt of the week 

Act as a data scientist and train a classification model to predict [target variable] based on [features] dataset. 

See you next week, 

Mukundan 

The Data Talent Pulse is brought to you by TeamEpic, a trusted global AI Talent provider. Learn more about us here.