Some tasks require understanding, not just knowing how to do. Tools can’t fill the gaps in understanding. For these tasks, time is better spent learning and understanding. No-code development is useful for building without understanding, but understanding is most critical when things fail. And things fail while building products, be they data products or otherwise.
Here the user switches from Cursor (automated coding) to Bubble (a no-code tool) to address the lack of understanding, not realizing that switching tools is solving the wrong problem.
We often make the same mistake in data science, especially in predictive modeling, where a new off-the-shelf library or method is treated as a prophet (pun intended), only to find out later that it was solving the wrong problem.