Yep. The challenge for OpenAI right now is how to drive cost-per-query down, particularly if they wanted to scale their capacity to something approaching Google scale.
Right now their cost per query is about 10-100x what it costs Google to perform a web search, of which Google performs 8 to 10 billion a day. So currently if they were to try to be on par with Google in terms of capacity their costs would be enormous. So driving those costs down is key.
Once that cost gets driven down, it's a whole new ballgame, and probably the first thing so far that could meaningfully challenge Google's core product.
There's going to also be extraordinary effort to tweak it, but it'll improve over time. That's not even taking into account business that will (and already are!) building on top of what OpenAI is doing, and tweak it heavily towards specific niches.