Well, I for one do not "trust" the accuracy of its output. It may very often be right, but it's not an expert system or truth maintenance system.
And especially if you're doing original research, trying to seek out primary sources and understand something that might not be in the more popular history texts, I would be even more dubious.
Well, you could get historical summaries from reference sources way before computers existed.
But the study of history is not about "summarizing" aspects and timelines from existing written histories.
It is more about reading those summaries, and even more so, reading more extended historical works and original sources too.
>Things like "describe the deities of the Akkadian empire, and which ones were preserved or influenced others in subsequent near-Eastern civilizations" and you get instant answers to what would be days or evens weeks of book-based research.
For personal purposes (e.g. not writing an academic paper on the subject) and for someone with a good reference library (or access to a good library), days at most. Often mere hours, as such things are often summarized in "popular history" or more specialized tomes. With the advented of the web and access to online academic libraries and papers, plus full-text search, even less.
And of course, ChatGPTs summary will often contain confident made-up BS.
Hostory is already dead before ChatGPT.
With the invention of smartphone and globally available internet, the volume of records' accumulation is in exponential growing rate.
For example if 911 happan in 2023, the total amount of videos and photos will exceeds 100TB, with different angles and perspectives. So all we know historians are not going to check all 100TB of records to find the truth. So what exactly are they trying to achieve?
They are going to read newspapers as what people did in the old days. But we know newspaper do not tell truth. Live streams do. Vlogs do. People burning inside the building leaving their last note on tiktok, that captures exactly what is happening, and historians will just ignore those.
If your “pursuit” is satisfied by reading catchy sounding blocks of text with no reason to believe its veracity, then yes, ChatGPT has rendered the study of history obsolete.
If by “study” and “history” you define those words in the way that most people would, then no, ChatGPT has not rendered the study of history obsolete.
I think your confusion here is about whether it has rendered your “study” obsolete to you. If I had no interest in studying history I would be inclined to agree with your point. Maybe the question is do you get any enjoyment or fulfillment out of studying history? Have you ever?
ChatGPT can't provide primary resources, and given its training material, can't put primary resources it has been given into a valid historical context. History as a discipline is safe.
What you're thinking of as history is just a collection of facts. And it'll lie about those.
7 comments
[ 3.3 ms ] story [ 28.3 ms ] threadAnd especially if you're doing original research, trying to seek out primary sources and understand something that might not be in the more popular history texts, I would be even more dubious.
I am more worried about people like you abandoning primary and secondary research under the misguided belief that chatgpt is correct.
Would you abandon your research and pursuits because some random person on reddit claimed something?
But the study of history is not about "summarizing" aspects and timelines from existing written histories.
It is more about reading those summaries, and even more so, reading more extended historical works and original sources too.
>Things like "describe the deities of the Akkadian empire, and which ones were preserved or influenced others in subsequent near-Eastern civilizations" and you get instant answers to what would be days or evens weeks of book-based research.
For personal purposes (e.g. not writing an academic paper on the subject) and for someone with a good reference library (or access to a good library), days at most. Often mere hours, as such things are often summarized in "popular history" or more specialized tomes. With the advented of the web and access to online academic libraries and papers, plus full-text search, even less.
And of course, ChatGPTs summary will often contain confident made-up BS.
For example if 911 happan in 2023, the total amount of videos and photos will exceeds 100TB, with different angles and perspectives. So all we know historians are not going to check all 100TB of records to find the truth. So what exactly are they trying to achieve?
They are going to read newspapers as what people did in the old days. But we know newspaper do not tell truth. Live streams do. Vlogs do. People burning inside the building leaving their last note on tiktok, that captures exactly what is happening, and historians will just ignore those.
If your “pursuit” is satisfied by reading catchy sounding blocks of text with no reason to believe its veracity, then yes, ChatGPT has rendered the study of history obsolete.
If by “study” and “history” you define those words in the way that most people would, then no, ChatGPT has not rendered the study of history obsolete.
I think your confusion here is about whether it has rendered your “study” obsolete to you. If I had no interest in studying history I would be inclined to agree with your point. Maybe the question is do you get any enjoyment or fulfillment out of studying history? Have you ever?
What you're thinking of as history is just a collection of facts. And it'll lie about those.