NOAA, one of the leading world institutions in climate science, predicted in October 2018 that the U.S. would have a warm winter and the West's drought would continue.
October 2018 prediction: https://www.noaa.gov/media-release/winter-outlook-favors-war...
March 2019 retroanalysis: https://www.climate.gov/news-features/blogs/enso/winter-outl...
They were way off. They can't predict the next six months with a high degree of accuracy, much less the next century.
And it has nothing to do with NOAA's inability to accurately predict six months out.
"More and bigger" are relative terms. Pegged to what date in history?
It has everything to do with NOAA's ability to predict 6 months out because uncertainty for predictions increases as a function of time. If you can't predict 6 months accurately, you have even less of a claim to predict 60 years accurately. That's how predictions work.
So if you see the number of records increasing over time (modulo new data sources), then something interesting is happening. Either the center point is moving, or things are getting more extreme. If you see more records of both ends (hot and cold, say), then you can say that yes, things are getting more extreme. If you see more records on both ends, but more records for heat, then you can say that things are getting both hotter and more extreme.
I don't have to have a quantifiable prediction of what the temperature will be in order to make that observation. If it were to be cast as a prediction, it might be something like "among weather stations that have existed for at least 30 years, there will be more temperature records set next year than there were 10 years ago".