4 ms·
I have a device in my pocket with 1 billion transistors switching on and off a billion times a second. I couldn't possibly check every single one of them, or al
by tambourine_man 10y ago
I have a device in my pocket with 1 billion transistors switching on and off a billion times a second. I couldn't possibly check every single one of them, or all the steps that went on its production.
I can, however, check the results it gives me and have a degree of confidence based on its accuracy and consistency.
We do just that with climate data and models.
People choose to challenge some aspects of science when it clashes with their interests or system of beliefs while completely relying on others when it's convenient. Even though the process by which scientific knowledge is produced is the very same. That's not only illogical, that's hypocrisy.
- alphapapa 10y agoIt's actually bizarre to see someone on HN conflate software and hardware. Your argument appears to be, "Hardware can be observed to operate properly, therefore software necessarily gives correct results." There are a variety of arguments to make here, from the flaws in individual models, to the secrecy of source code, manipulation of data, politics, historical examples of incorrect scientific conclusions, and the general fallibility of humans (who perform the science). But bother with any of those when you are claiming that, if software exists which ostensibly intends to perform a task, it must necessarily do so correctly. And then you accuse others of illogic and selective acceptance of science. That's just...bizarre. In the context of discussing belief vs. knowledge, that smacks of what one might call a "true believer."
- tambourine_man 10y agoThe argument is that software can be used to check hardware, and, more broadly, that one does not need to execute every step of a process in order to have confidence in its results.
- alphapapa 10y ago> software can be used to check hardware Sorry, but how is this point relevant to the discussion? > more broadly, that one does not need to execute every step of a process in order to have confidence in its results. That is broad, indeed. I should hope that, before reaching conclusions and making claims and calls to action that would result in the poverty of billions, one would take pains to verify every step of the process used to reach those conclusions. Of course, what constitutes "verification" is a matter of opinion. To some, the models are sufficient. To others, they are grossly insufficient and have not proven accurate looking forward nor backward. It is a mystery why they continue to be asserted as proof.