I'm a former Google engineer who founded an AI startup. Here are the AI tech skills to learn and the tech roles AI will replace.

• AI will radically alter how computer scientists do their jobs. 
• Ex-Google engineer and AI startup founder Zach Smith learned this first hand.  
• AI will overtake many tasks, and some roles, he says, leaving people free for more creative work.

This as-told-to essay is based on a conversation with Zach Smith, founder of Nova AI. It has been edited for length and clarity.

When you think of the sort of people who land high-paying software engineering jobs in big tech companies, you might imagine teenage tech geniuses who got computer science degrees at top universities. And it's true that many of them are.

But the path to my dream career – founder of a newly launched startup called Nova AI – wasn't straightforward or easy for me. And now I want to share what I can clearly see is the future of jobs in tech: Some functions are going to go away, replaced with AI. And the people who do them should prepare themselves to be shifted towards different, more creative tasks. 

From college drop out to a master's degree and job at Google

When I was 19 years old, I dropped out of college because I was studying finance and knew it wasn't for me. I spent a couple of years trying a bunch of stuff: real estate, credit finance, all having no degree and none of it was working for me.

I knew I had to go back to school and opted for computer engineering because I was a builder, meaning I always liked making things. Once I decided on engineering, I crafted a future resume for myself that articulated both current and future milestones. This gave me an ideal state to work towards that would, in the near future, be all-encompassing of the skills I already had and the skills I would acquire. I looked at the best courses that were free or affordable in conjunction with a formal Bachelor's route.

This path of self-education, in addition to schooling, was how I secured my first job as a software programmer at a large company prior to finishing my degree. Once I chose my path, I stayed in school until I earned a master's degree in computer science from Georgia Institute of Technology at age 26. I loved that program. It was fully online but cost effective and I could also continue working while I did my masters.

One job led to another until I obtained my graduate degree and landed at Google working on cloud-related teams like one that helps Google cloud customers use a lot of automation technology.

I don't want to make it sound like I just skipped my way from one job to another. Technical job interviews are difficult and I found resources and studied for each one in order to do well. (One resource I recommend is Cracking the Coding Interview by Gayle McDowell.) 

Earlier this year, I left Google and launched Nova AI. We are building a startup that uses AI to automate repetitive software engineering chores, particularly in quality assurance (QA) work, helping free up engineers time to work on more creative tasks, like building new products and features.

So when I say that embarking on a career in software development can be both exciting and overwhelming, it's because I've been there. 

Now, in my work as a startup founder, there's a new area to consider for people with careers in tech: You will need to be relevant in the age of AI.

AI impacts QA engineering, technical writing

In truth, staying relevant in the age of AI is going to impact a lot more people than just computer engineers. For instance, I know people working on AI bots that replace fast food workers who take drive-through orders.

But in the tech world, "AI is like having an orchestra of tireless worker bees, buzzing away at the repetitive tasks, while you, the master beekeeper, focus on crafting the sweetest honey of innovation," That's what Chat GPT 4 told me when I asked it to describe AI's role.

For instance, Nova AI is building features that automate some of the most repetitive tasks that quality assurance engineers currently do such as user acceptance testing. That's when the QA engineer goes through a product manually clicking on everything to test all product features.

While I don't think that QA engineers are going to be entirely replaced, all of the manual functions they do will be automated. QA engineers will become the ones that run and train these AI automation tools. Each QA engineer in the future will be far more productive, meaning companies won't need to hire as many, but also that people in these roles will be freed to work on more creative tasks. User experience (UX) design is the type of creative role that AI can't replace.

Another function I believe AI will automate is technical documentation. AI will be able to look at what the code does and then generate the documentation that describes it. We're using some of those tools here at Nova AI.

The third function that is going away is code maintenance. Code gets old, libraries change, everything needs to be updated to the newest versions, newest patches. Currently many engineers spend a chunk of their time doing this kind of work. 

In fact any function that relies on stitching – do a task and take the output of that and do the next thing – AI will do all of that in the future. 

On the other hand, AI is also going to create a lot of jobs: data science, machine learning engineering (ML), AI ethics and design will be new functions in high demand. Within AI there will be many specialities: auditory ML stack (dealing with sounds), computer vision, robotics.

So yes, AI will change the tech jobs landscape from what it is today. But despite the doomsayers, humans are creative. AI only knows what it's been trained on. The next generation of jobs will be more creative and less mundane. People won't be replaced. They will be shifted.