Working With AI
Bruce Randall on Getting Real Work Out of AI Tools
The AI strategist and technology seller shares how he prompts, checks and builds with AI, and why critical thinking decides how much value anyone gets from it.
A Discussion With Bruce Randall
Video from A Guy With AI.
Generated from the canonical interview transcript and validated against source data by Guests on Air.
How can professionals get better results from AI tools?
Bruce Randall says better results come from asking precise questions, following up by asking what else the tool would suggest, and checking every output for accuracy and sources. He pairs AI with his own critical thinking, uses tools such as Perplexity and Lovable, and urges people to start with one practical use rather than fear the technology.
Bruce Randall has spent years selling technology, including AI, and has added a data science course at MIT and a hands-on app-building course to that experience. On the A Guy With AI podcast, the host asked him how he actually works with the tools day to day, where they help, where they mislead, and what he tells people who are still wary of them.
His answers were practical. Randall treats AI as a partner that rewards good questions, needs its work checked, and often returns more than it was asked for. He also sees it spreading quickly into everyday business, citing a small local contractor that already uses an AI voice agent to take customer calls, a sign that almost any company can now afford it.
The thread running through the conversation is critical thinking. In Randall's view, the people and businesses that gain most from AI are those who bring their own judgment to every output, keep asking questions, and start using it now, rather than those who hand over their thinking or wait until competitors have pulled ahead.
Key takeaways
- After a good answer, ask AI what else it would suggest; the follow-up often surfaces ideas you missed.
- Check every AI output for accuracy and sources, because tools sometimes add material nobody asked for.
- Randall went from idea to a working app in eight weeks with no earlier app-building experience.
- Pair your own critical thinking with AI; he argues that combination produces the best answers.
1. Ask better questions, then ask one more
Randall compares prompting to using a search engine. People who describe what they want loosely often need several attempts before the answer fits, and some people simply ask better questions than others. With AI, he says, you have to ask the right question in the right way, then fine-tune as you go until the output matches what you actually need.
His favourite habit comes after a good answer arrives. He asks the tool whether there is anything else it could create that would benefit him, or anything he did not think of. The results often surprise him by surfacing ideas he had not considered. He calls these gems, and says they sometimes send a project in a new and better direction.
He applies the same thinking to how information is processed. His data science study showed him that the formulas chosen in the back end move results closer to or further from the goal. That is why he reads research with an eye on who funded it, and why he values having control over how AI processes data inside a workflow.
2. Check the work and choose the right tools
Speed is the obvious benefit, but Randall warns that AI can slip in material that was never requested and that he cannot always explain. Anyone writing a paper with AI, he says, has to check that everything has sources and is accurate. The tool works faster and sometimes more accurately, but responsibility for the final result stays with the person using it.
His own toolkit reflects that balance. He likes Perplexity because it draws on other AI systems, combines the information and adds footnotes so he can verify claims. For building apps he moved his whole workflow into Lovable, which returned a working flow along with a designed web page and a better web address than the one he had chosen himself.
He also notes how far coding assistants have come. Randall knows people who build apps entirely with AI and never write the code themselves, which means you no longer need to be a programmer to ship something useful. For him, that is the clearest sign of how much the tools can enable people who are willing to learn them.
3. Critical thinking is the multiplier
Randall cites a research article arguing that some people who use AI stop thinking while others get sharper, and he draws a wider lesson from it without fully agreeing. Critical thinking, he says, became a buzzword because it was partly lost. Combined with AI, it gives you the best of both worlds; without it, answers take longer and miss the mark.
He is applying that lesson to his own book, The AI Human Paradox, which he is rewriting for a broader audience. His first draft proved too dense for most readers, so he is using his own judgment, with some help from AI on specific topics, to make the ideas friendlier. He treats the rewrite as an exercise in clearer communication.
His closing advice was direct: do not be afraid of AI. He had never built an app before taking a course, yet went from idea to a working app in eight weeks. If you question something, try it, he says, because failure is one step closer to success. For business owners, his advice is to find one use that helps now and build from there.
