Episode summary
Why AI Is Still in the ‘Toy Phase’ (and What Breaks It Out) | Brian O'Malley
How I Invest With David Weisburd
· · 9 min · Hosts: David Weisburd, Curtis Pierce · Guests: Brian O'Malley
Summary
Brian O'Malley argues AI is starting to move beyond the early “toy phase,” but it still has a long way to go before people trust it with meaningful decisions and real dollars. He says the earliest ChatGPT use cases were novelty-driven, while the more interesting shift is toward higher-value queries and monetizable behavior, especially where AI can influence purchases or advertising outcomes.
The main constraint, in his view, is not raw intelligence but execution. AI becomes much more useful when it can connect digital recommendations to offline outcomes like booking travel, coordinating events, or handling paperwork. O'Malley points to travel and group-event platforms as examples where AI can do much of the work, but a human still needs to validate the final step because hallucinations remain a real risk.
He also draws a line between tasks where 95% accuracy is effectively enough and tasks like taxes or legal filings where the last few errors matter a lot. For O'Malley, the near-term winners are companies that use AI to automate 95% of the process while keeping a human in the loop for trust, empathy, and economic efficiency.
Key takeaways
- O'Malley says AI is still partly in a toy phase because users are not yet comfortable giving it trust, preferences, and hard-earned cash.
- He believes AI gets more valuable when it bridges digital answers to offline execution, not just digital spending.
- He cites travel and group-event businesses as early agent use cases, but says a human is still needed for the final mile.
- O'Malley says 95% accuracy is good enough for emails and blog posts, but not for taxes or other high-stakes tasks.
- He argues the economics of training and deploying AI will matter more once gross margins and profitability become central.
Notable quotes
“Been a couple weeks since we chatted. And I would even argue that we're we're starting to uh move out of that”
“at the end of the day, it takes a couple things for AI to really move out of this toy phase”
“So how does AI not just give me a recommendation but actually go ahead and execute that recommendation with my dollars”
“when you start talking about really important events, you only get to take a vacation a couple times a year”
“if you're going through a life event where you need social security disability, you want to feel like there's a human helping you on the other side”
How the episode opens
When we last chatted, you said that we were in the toy phase of AI. What did you mean by that?
Been a couple weeks since we chatted. And I would even argue that we're we're starting to uh move out of that. So much changing so quickly.
But if you think about the early cases when chat GPT first came out, I remember when the first thing I did would, you know, be write like a silly song and have it sung in a pirate voice or something like that. There was a lot of this playing around because it was so different than what you would experience with any other product online. But from there, people started moving into more serious queries.
Uh there's obviously a lot of usage in around education, maybe less monetizable, but people using AI for more serious endeavors. You now have more people asking AI questions about more monetizable events, which you could see from an advertising platform would have real intrinsic value going back to these platform model companies. But at the end of the day, it takes a couple things for AI to really move out of this toy phase.
And I look at it in two ways. The first of which is trust from the user to be able to share more about both their existing preferences as well as their existing third party profiles. So how does AI not just give me a recommendation but actually go ahead and execute that recommendation with my dollars.
So I'm trusting the system with my hardearned hardearned cash. But also the second thing is that 85% of purchases still take place in the offline environment. And so when you think about an Uber, when you think about a Door Dash and Airbnb, those tools are ultimately connecting the digital and the physical world in a clever way.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai trust
- ai execution
- ai novelty phase
- physical world integration
- real-world adoption
Watch the episode
Watch “Why AI Is Still in the ‘Toy Phase’ (and What Breaks It Out) | Brian O'Malley” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from How I Invest With David Weisburd
-
Hunter Walk: Why Venture is Getting Harder, Picking Great Founders, and Investing Against Consensus -
True Anomaly’s Even Rogers on Space Warfare, China’s Military Advantage, and the Defense Tech Boom -
Finding the Next Anthropic, Quantifying Venture Capital, and the Science of Picking Winners -
Why 95% of AI Projects Fail, Anthropic still at $2T?, and Why AI Is Still in the First Inning -
Peter Thiel's 27-Year Partner on the AI Bubble, the Venture Capital Crisis, and the PayPal Mafia -
Why Venture Is Eating Public Markets, Picking Great VCs, and the AI Boom w/Brijesh Jeevarathnam
How I Invest With David Weisburd, this episode and its artwork belong to David Weisburd. Octagon is not affiliated with the show; this page is an independent summary.
