ChatGPT obsoleting Encyclopaedia and Textbooks?

Doesn't ChatGPT rely on the information contained within encyclopedias and textbooks? I've heard there is a breakdown in fidelity of information when ChatGPT begins to use its own output as a reference source. Its answers degrade into gibberish.

Encyclopedia and textbook are valuable as an enclosed and marked source of accurate information from which AI derives answers to questions. In that sense they can't become obsolete.

Reply to Nils Loc

I am not sure on the exact information on ChatGPT myself, because I just discovered it a few days ago, and tried it. Yes, it would be relying on the information contained within encylopaedia and textbooks, but it would have more uptodate information due to the ability of frequent update, whereas the printed books get no update once published. The readers must buy the new editions, if the publishers decided bring out new editions with the updated content in the books, and if they want to buy the new edition books (I don't tend to unless there is need for me to buy the new edition copies).

I was on the impression, that ChatGPT might spew out gibberish at times, especially if the input content is more complicated ones. But for the simple concept searches might be handy resource to have for quick reference. With the A.I. technology set to improve in the future, it will get better I suppose.

Quoting Corvus

Will the textbooks and Encyclopaedia will still be in demand?

I voted the Encyclopaedias and text books will still be in demand. ChatGPT is a good invention, but it is just a tool. Furthermore, the answers are not reliable enough, whilst an Encyclopaedia is written by authors who are experts in the ground. On the other hand, if I am not wrong, I guess that the functionality of ChatGPT is the elaboration of replies thanks to the data that it collects from Encyclopaedias.

Quoting Corvus

Will the textbooks and Encyclopaedia will still be in demand? (6 votes)

Yes, they will still be in demand. The good old paper books will always be with us, it looks like.

Textbooks and encyclopedias are not influenced by the student. The data in them is collected, organized, elucidated, presented and footnoted in an integrated body of information, rather than snippets to answer limited, possibly disorganized or even incoherent questions.

Reply to Vera Mont Reply to Vera Mont
I will be keeping all my old textbooks, Encyclopaedia and the philosophy books.
At one point, I was thinking of getting rid of the old paper books, but they will come handy for reading, referencing and studying.

Which, of course, is not to say that one shouldn't make use of on-line information sources (including dictionaries and encyclopedias), especially for updating statistics and recent scientific advances.
I prefer government departments, university and professional organizations, but there are solidly researched and well organized websites dedicated to history, climate change, engineering - all kinds of specialty subjects. It's a big, ad-infested goldmine of knowledge - why restrict yourself to a chatty robot?

Quoting Vera Mont

why restrict yourself to a chatty robot?

I wouldn't be interested in chatting to the robots at all. I prefer chatting to the real humans for sure. :)

Quoting Vera Mont

It's a big, ad-infested goldmine of knowledge - why restrict yourself to a chatty robot?

Well, the chatty robot can do a lot of the searching for you and provide a quick answer.

Quoting jgill

Well, the chatty robot can do a lot of the searching for you and provide a quick answer.

How do you know the quick answer is the correct answer?
I suppose your search criteria will depend on your reason for seeking the information. If you're a student, a quick answer to a specific question might fill in one tiny gap - but it won't help much in understanding the subject or retaining enough to pass exams.
If you're looking to score a point in a forum discussion, a quick answer is great - so long as your interlocutor accepts your source as authoritative. While the robot's sources may well be impeccable, you can't prove it.
If you're just doing research for a story, or can't recall something you knew, or need a precise measurement for something of which you know the approximate dimension, a quick check may be all you need.

Quoting Vera Mont

While the robot's sources may well be impeccable, you can't prove it.

True. But BingAI lists the internet sources, so you can go directly to them for verification. Though some might be shaky, as any search might show. What can you believe on the internet? On Wikipedia, for example, I have found advanced mathematics pages to be very accurate, but elementary pages not necessarily.

Quoting jgill

On Wikipedia, for example, I have found advanced mathematics pages to be very accurate, but elementary pages not necessarily.

This is probably why wiki is not recognized by academic institutions. I've been on a couple of forums where it wasn't accepted, either. It's easy; it's handy; it makes you feel guilty about once a month when you don't donate, but its best feature is the links to original sources.

Few year ago, I used to distrust information from Wiki or any online source, and the people who copy & paste the contents from the online sources into the forum posts. But now, I am more open minded on online information sources for their accuracy and quality.

I still tend to pass the copy and pasted contents from online sources in the posts unless it is vitally relevant to the points or topic.

There are many reliable sources on line. Of course, for forum posts, the level of rigor is much lower than, say, a university paper, but some interlocutors are too knowledgeable to fob off with estimate or dubious statistics. Some subjects, too, are more demanding than others. If I want a general idea of what conditions were like in a historical period for the setting of a novel, I prefer sources with lots of illustration, so I can form a picture, but then I have to go more detailed accounts of correct language, attire, armaments, furnishings or whatever. If I'm trying to convince someone that vaccinations are effective, I'd better not cite incorrect numbers.

Breaking news: NY Times sues OpenAI for copyright infringement

The New York Times is suing Microsoft and OpenAI, the creator of ChatGPT, claiming millions of its news articles have been misused by the tech companies to train their AI-powered chatbots.

It's the first time one of America's big traditional media companies has taken on the new technology in court. And it sets up a showdown over the increasingly contentious use of copyrighted content to fuel artificial intelligence software.

The legal complaint, which demands a jury trial in a New York district court, says the bots' creators have refused to recognise copyright protections afforded by legislation and the US Constitution. It says the bots, including those incorporated into Microsoft products like its Bing search engine, have repurposed the Times's content to compete with it.

This image shows an instance of the alleged copyright violation, with the red text identical to that in the NY Times original




More from NY Times (gift link) and Australian Broadcasting Commission

Quoting Corvus

Will the textbooks and Encyclopaedia will still be in demand?

The print textbooks will remain a primary source of specialized information. The print book encyclopaedia are dead ducks. Wikipedia is the best thing on the internet IMO.

I suppose chat bots can be useful, but when a chat bot gives us almost instant access to answers, we learn nothing from the process of getting it.

By slowly wading through encyclopaedias, text books, online articles etc. we accumulate knowledge.

Quoting jgill

The print book encyclopaedia are dead ducks.

Why do you believe so?

Quoting Corvus

Why do you believe so?

They can't be continually updated, like Wikipedia. They cost $.

True, some knowledge doesn't change, but Wikipedia has that too.

Quoting jgill

They can't be continually updated, like Wikipedia. They cost $.

I own an Encyclopaedia of Philosophy (MacMillan) published in 1960s, and it is still very useful, readable and feel up-to-date. I don't see a need for update of them. But suppose, for the other subjects, it could be different situation with updating.

Quoting Corvus

I own an Encyclopaedia of Philosophy (MacMillan) published in 1960s, and it is still very useful

Yes, I can understand philosophy and history retaining their value over years. Science and math change much more.

Quoting Corvus

With the advent of A.I. and the use of ChatGPT getting popular, I wonder if all the Encyclopaedia and Textbooks become obsolete. Would it be the case, or the textbooks and Encyclopaedia will still be in demand?

I did not vote. When you say with the "advent of AI and the use of ChatGPT" --are you saying we are also doing away with the editors, publishers, scholars, and reviews of the references and citations? Because those were what it took to create those books.

The scholars are researchers who are the specialists of a given subject matter.

So, I don't understand the question. And are you also including in your question the copyrights? Is authorship also obsolete?

Quoting jgill

Yes, I can understand philosophy and history retaining their value over years. Science and math change much more.

Yes, It was my best buy from eBay last year costing around $50 including the delivery of the 8 volume TEP set.

Quoting L'éléphant

are you saying we are also doing away with the editors, publishers, scholars, and reviews of the references and citations? Because those were what it took to create those books.

In case of the online information such as from WiKi or ChatGPT, the editors, publishers, scholars ..etc source information can be unknown or vague. And also the quality and accuracy of the information could be a bit suspicious too.

I prefer relying on the information from the traditional printed books and articles for the clearer information of the source, editors, writers and publishers.

Quoting L'éléphant

So, I don't understand the question. And are you also including in your question the copyrights? Is authorship also obsolete?

I would think the copyrights issue will always be with us. If you wrote something, and published it, then you wouldn't want someone quoting them without acknowledging your authorship or asking for your permission to quote or use them for their uses, would you?

Quoting Corvus

In case of the online information such as from WiKi or ChatGPT, the editors, publishers, scholars ..etc source information can be unknown or vague. And also the quality and accuracy of the information could be a bit suspicious too.

I prefer relying on the information from the traditional printed books and articles for the clearer information of the source, editors, writers and publishers.

It's more than a preference, but yes, I voted that the printed books are very relevant. There are legal properties attached to the physical copies of a book -- it is a tangible property which is regulated by the distribution, copyrights, printing, plagiarism laws. Out-of-print books could be re-printed. The years (20xx) and number of copies printed become its valuable properties.