A unique background

Tristan is a South Africa-based QA with a degree in psychology and communications. He started his career as a music journalist, where he honed his writing and editing skills. After transitioning to trade publications, he worked with specialized professionals like engineers, events professionals, logisticians, among others.

Mindrift Team: Do you think your degree in psychology ties into AI training well?

Tristan: It helps when you’re dealing with people, especially in this capacity where it's not face to face. You've got to think of ways to effectively communicate while remaining direct, empathetic, and non confrontational. I think understanding people and what their pain points may be and the complexities of interpersonal communication definitely helps on the human side of AI.

Mindrift Team: What inspired you to get into AI freelancing?

Tristan: It was a serendipitous alignment of conditions, I suppose. The company I was working for had to shut down and while I was job hunting, I came across a post for an AI Tutor - Editor. I had a look through the specifications and I realized, well, it's actually what I've been doing already, just in a more forward-thinking, future-proof context. 

The Mindrift experience

When Tristan started working on his first project in 2023, he felt like it was a “dream alignment.” Although he works with other clients outside of Mindrift, he feels like the flexibility the platform provides allows him to seamlessly integrate it into his schedule. 

Mindrift Team: Why did you decide to collaborate with Mindrift specifically?

Tristan: A lot of the AI-related posts are very wishy-washy and seem a bit undefined. With Mindrift, it seemed like a much more personal human experience. I think a lot of the competition today doesn't really take that approach. Everything's automated—you're really just a number in the system. You can log a support ticket and never speak to a human being. And from my first interactions with Mindrift, it was obvious that you'd actually be able to create human connections while working on building machine intelligence.

QAs: Custodians of quality

When chatting about the QA role, Tristan refers to it as “custodians of quality” dedicated to ensuring what's being fed into the models makes sense, meets the requirements, and will lead to model improvements. 

Mindrift Team: Can you explain what QA in AI means for beginners? 

Tristan: So, the top-level view of it is that QA serves as the quality control for all the content that's produced by writers, annotators, or whoever is creating prompts and/or responses. What we typically do is look at the expectations, particularly the formal and technical specifications, of a project. 

We look at the client or project's goals and ensure that everything submitted by our writers or experts meets these specifications and is up to standard, in terms of natural language. So, we are very much language and content editors in the old school sense, but working in a new context. 

Mindrift Team: What’s the most surprising, yet interesting, part of the role?

Tristan: There’s a very interesting and enriching aspect of this job: working on different projects and getting insight into how people from different fields interact and interpret the world. It’s been kind of eye opening.

Mindrift Team: How do you ensure high quality outcomes on the projects?

Tristan: I think it starts off with effective communication around the project goal and requirements, and holding people accountable when it comes to following those. And then there's obviously the aspect of actual hands-on editing. We also measure quality metrics using certain data analytics tools, and that helps us intervene with highly productive individuals who may need some assistance. So we've got sort of a multi-level series of checks and balances to ensure that everyone's doing their part to maintain quality. 

All things AI

We love hearing our freelancers’ takes on what AI means to them, and how they envision its future impact on society. 

Mindrift Team: Let’s talk about AI applications within your domain and some real-world applications that you think are exciting for AI.

Tristan: I've worked a bit in advertising and broader communications, and I think AI is a very valuable tool for inspiration. There's definitely amazing progress being made in visual models. I think it’s an incredibly useful tool for individuals working in a variety of spaces, like events and design. 

I also think in terms of knowledge accumulation, learning things—it’s a sort of self-education space. Everyone can benefit from that, and that's why it's very important that the models are fed good information and learn appropriately. But we will reach a point where the information, knowledge, and decision-making of machines is better than humans, and that's an interesting quandary.

Advice from the experts

When discussing QA roles, Tristan believes that although there’s room for people who simply enjoy writing and editing, domain experts are most in demand. He believes “the future of AI training is very much narrowing in on complex professional or niche fields.” 

Mindrift Team: Do you have any advice for some of the newer freelancers on Mindrift? 

Tristan: Because it’s a freelance position, you need to be actively engaged with other clients. You simply cannot put all your eggs in one single basket, because realistically, it’s not a full-time gig. And I think Mindrift really does the most it can to provide people with as many opportunities for work as possible. So, I would say manage your expectations because projects are very much dependent on the clients’ requirements. But I think it's a privilege to be able to work on projects like this.

Mindrift Team: Describe the type of person you think a QA role might be suited for?

Tristan: I think the profile of the ideal AI tutor is very broad. What you do need, obviously, is a good grasp of the language that you’re writing in, but you also need to be receptive to feedback, somewhat dynamic, and not resistant to change. Things evolve very rapidly in this space. 

If this sounds like you, and you have professional experience in any domain, come join us at Mindrift! Mindrift is a platform that curates a talent pool of experts and connects them with cutting-edge AI projects from the world’s leading companies. 

Work on interesting AI projects on your own time, add new experiences to your resume, and get paid for your contribution. Check out our open roles and talent pools to get started.