Bryce Maddock, CEO at TaskUs, shares his views about the new jobs, roles, and challenges, that are emerging in the BPO industry as AI becomes more prolific.
---
Q: What AI-related jobs do you see emerging in the BPO industry?
A: We're seeing growth in what we call “AI Services.” For TaskUs, this has historically focused on computer vision for autonomous vehicles, but the industry has evolved to where more complex LLM models are prolific.
More recently, services have been about prompt-and-response writing. Through our global gig worker platform, TaskVerse, for example, we recruit people with Master's Degrees and PhDs in a variety of subjects that enable them to write accurate responses to complex questions that are used for training LLMs.
We've also seen a significant increase in work related to LLM safety. For instance, our red teams provoke models to say or do inappropriate things. LLM developers then add guardrails to prevent the model from generating this type of content once it’s made public.
Q: What other AI-related ethical concerns and risks should BPOs plan for?
A: When customers talk to AI agents that feel human-like, they’re more likely to follow that information blindly. This trust is an issue when an AI tool hallucinates. The results might be an inappropriate refund or a chatbot taking actions that are against the company's policies.
There's also the ethical question of letting the customer know when they’re talking to an AI agent. Customers often ask the AI to speak with a person. We believe you should be transparent and provide customers with that option. Human supervision of AI is the best solution.
Q: How is the balance between automation and human interaction evolving?
A: Technology is ideal for refunds, learning company policies and figuring out how something works. However, human interaction is critical in a range of other scenarios.
For instance, when implementing complex code on a website, it’s more efficient to get step-by-step guidance from a person who can handle ambiguity better than technology.
Similarly, for large purchases, speaking with an advisor or meeting face-to-face makes sense. Think about the real estate market. Despite the potential for full automation, almost all home sales in the US still involve an agent.
These instances suggest a dual mandate: simple tasks and repetitive transactions will likely be automated, while complex interactions will require increased investment in human talent.
Q: As efficient AI experiences become a customer expectation, how will TaskUs keep human teams relevant?
A: We believe in augmenting our employees' skills by supporting them with smart technology. We’ve developed a set of AI tools that enable our teammates to create more efficient, higher-quality customer experiences.
Typically, once a customer of ours has successfully automated simple, repeatable customer contacts, they want to invest more in higher-quality and white-glove service or potentially upselling.
To that end, we're rolling out training that's focused on the jobs of the future, preparing our teammates to take on Tier-3 technical work or turning them into excellent sales advisors.
The future will surely be more tech-led, but the customers of the future will still expect to connect with human experts.
---
Follow Bryce on LinkedIn ---> here
Don't miss future Q&As. Join our mailing list ---> here
Comments