The part you’re missing is that it’s not “the same data” but multiple data sources showing the same trend. Ice cores, stalagmites, lake sediment, tree rings, sea shells and corals, microfossils, etc. are all measuring different things - when those show consistent results it’s increasingly unlikely that this is coincidental.
Wikipedia has a good overview of the major categories which is well worth reading:
Those cover many separate fields so it’s also unlikely that this is because they all happen to be making the same mistakes. Revealing such a flaw would be a career-making move for an academic.
If you thought that science was certain - well, that is just an error on your part.
Richard Feynman
Feynman wrote a lot about the problems of science and scientific group think. Scientists are people. They have egos, careers, and families to feed like everyone else. They actually tend to resist outside results from their field consensus because they have papers and grants and reputations built on prior findings, as well as a healthy resistance to the newest hot theory that breaks everything because it is simply incorrect and they don't want top go through the trouble of really examining it.
Without the ability to check the techniques of paleoclimatology we have a have a science with extremely wide confidence bands on its estimates.
I think Hansen is absolutely right about Thorium nuclear as the carbon free energy pathway of choice. It certainly doesn't hurt anything to make that happen then is worse than catastrophic anthropogenic climate change. However, the choices made under alarmist predictions from theoretical models must be avoided.
I would strongly suggest learning what is common in the field - whoever is originating the talking points you’re repeating is doing you a real disservice.
One of the first questions to learn about is why you incorrectly believe there’s no way to validate techniques, and especially whether it might be the case that there are so many different lines of evidence is because scientists looked for independent sources of data so they could do exactly that.
My point is that we have no actual way to confirm how correct these models are because we can't check them.