Science requires systematic observation, measurement and usually variation (often experimentally controlled); and, usually, iterations.

One datapoint outside such a system is not science.

You can't even necessarily just insert a new datapoint into a pre-existing scientific sytem. The system itself may need to be adjusted, for example to test and account for biases that often occur due to how observations are made.

Not to my mind, science requires a testable hypothesis and evidence. I would argue that merely refuting someone else's hypothesis without providing a new one doesn't meet the bar of doing science.

Speech-to-text set to the wrong language or something?