Well, for starters, the graph lopped off the first twenty years from the data set so that it could "start" from the massive peak in the 20s and 1930. The top reply to your original tweet is a retweet of two videos that rebut the graph.[0]
It's further skewed by the use of decadal averages, which hid the fact that the greatest peaks included deaths that were either the direct result of--or greatly worsened by, conflict and/or a handful of specific policy decisions--food production failures during and conflicts such as the Zhili-Anhui War in 1920-21[1], floods that occurred during the Chinese civil war which dramatically worsened responses and recovery, the Holodomor in Ukraine in 1932-33 and the Soviet famine of 1930-33 more broadly, the 1938 Yellow River Flood[2] following the intentional destruction of dikes in an attempt to slow the Japanese Army's advance, World War 2 more broadly in the 40s where you had both the food production interruptions of war on a massive scale and explicit acts of mass starvation, the Great Chinese Famine in 1959-61 which is considered to be one of the largest man-made disaster in history,[3] etc.
The graph falsely suggests that we've we've somehow stumbled upon a viable adaptation strategy that makes climate change nothing to worry about. Since 1900, we've seen massive medical advancements, improved early warning systems for at least some types of disasters, transportation networks and technology that helps move people away from disaster zones both before some disasters and in their aftermath, the ability to rapidly move large amounts of food to disaster areas, and more.
Those are all great achievements, but the largest factor in the decline your chart suggests (albeit through data misrepresentation) is the fact that we don't have massive conflicts on the scale we saw in the first half of the 20th century, genocidal dictators looking to quickly wipe out millions of people through starvation, or political ideology driving inane agricultural policies that killed tens of millions of people because the autocratic dictators of some of the most populous nations on Earth read some pseudoscientific drivel (Lysenko and others managed to inspire not only the Soviet Famine in the 30s but also the Great Chinese Famine in the late 50s) and decided it sounded pretty ideologically reliable. We still have conflict and famine, but nothing on the same scale.
Trying to take that and spin it as climate adaptation is, well, absurd. Even by climate skeptic standards, that argument's a real stinker.
0. https://twitter.com/TheDisproof/status/1633492932484374530
1. https://disasterhistory.org/north-china-famine-1920-21
2. https://en.wikipedia.org/wiki/1938_Yellow_River_flood
3. https://en.wikipedia.org/wiki/Great_Chinese_Famine