Nearly two years ago I wrote an article titled "Whose Jobs will be Taken by Computers?" where I made a comment suggesting that Buzzfeed would let AI do the work of their human staff once it they realized that much of the content they produce could be made by AI. A few days later I decided to jokingly write a simple program that could generate a Buzzfeed style quiz and it has become a running joke on my site ever since. Well the other day my prediction came true and Buzzfeed announced they are going to start using AI for content creation, so if anyone from Buzzfeed is reading this I'd be happy to sell my quiz generator.
But joking aside my prediction coming true is not something that we should celebrate, not just because a bunch of Buzzfeed employees will soon be looking for a new job in a market where they've been deemed obsolete but because this marks the next step in the decline of, for lack of a better term, quality internet content. For years now humans have not been the target audience of the internet, Google has been and for the most part always will be. Search engine optimization means that people writing online articles are no longer writing to entertain humans but to please a machine, imagine how much better a machine can be at gaining favor with another machine than a human could ever be able to. Then if we stack on top of that the fact that the first few paragraphs of articles that a search engine would give us are filled with fluff to boost SEO and force us to scroll down so that we see at least one more advertisement (this is why recipe sites always put a story before the recipe) we see that the internet is becoming more and more useless and inconvenient for finding information, the thing which is supposed to make it most worthwhile.
So how do we fix this problem? Using alternate search engines is one idea. Marginalia is an interesting one that I recently discovered, it recently reached the milestone 100,000,000 indexed documents but due to hardware limitations it doesn't make sense for its creator to push for quantity, in the post announcing this achievement it was said that "Focus should instead be on improving the quality of what is indexed, on making it better, faster, more relevant." This is a noble goal, filter out the junk and give people relevant quality information, is there a good way to do that? Well if somebody had already found one I wouldn't have to write this article.
Wiby is another niche search engine that focuses on indexing simple non-commercial sites like mine. Most of what it has indexed are human submitted pages which makes it a fun tool for discovering sites you wouldn't find elsewhere but rather terrible at reliably finding specific information, which is the whole point of a search engine. If today you used Wiby to search the term "Ford Lightning" the first and only relevant result that comes up is my article titled "The Ford F150 Lightning" and when I say that is a relevant result I am using the term quite loosely because that article is more of a rant on electric vehicles than an informational piece on Ford's electric truck. There are a few results relevant to other Fords and if you scroll down even a little bit you find my cryptocurrency article which has nothing to do with electric trucks or even the original truck that the name Ford Lightning came from. Marginalia doesn't do much better on the Ford Lightning test, these search engines are cool in theory but in practice you are better off reading whatever AI generated content Google feeds you when you need specific information or recommendations, occasionally you get lucky with them and stumble upon an artisanal webpage but the odds of it having precisely what you need are slim.
There are of course other ideas people across the internet have had to make it more useful for storing and finding information. Personal wikis are an interesting idea and have the potential to be of great use but mostly only to their creators. Saving everything is of course another option for people seeking to keep hold of the stuff on the internet they find useful but then indexing becomes a problem again. Exclusively using quality specialized websites, like based.cooking, is probably your best use of the internet for information gathering but sites like these are hard to find for many topics if they exist at all. Where am I supposed to find a similar online resource for beekeeping advice, car repair, or carpentry? And would such a site even be that useful for those kinds of topics?
Because of the recent advances of AI and the degradation of the internet many people have been talking about this same topic and many of these same solutions along with others but all of the people I've seen talking about this problem have neglected the most time-tested solution to storing and accessing information, books. If you really want to learn something buy a book about it.
This was the topic of one of my earliest blog posts (back before I was all that great at writing these). I went to school for computer science and while I was there I didn't learn to program by attending class, I learned it by reading books (which the classes were based off of). I'd read the textbook then when I'd get to a part in my assignments that I had trouble with I wouldn't look for the solution online, I'd reference the book. I kept those books throughout my time as a student and I'd go back to them from time to time because looking there was better than looking online. My dad and I learned beekeeping by reading a book and when we were starting out we'd reference that book often and my dad has bought more beekeeping books since then, I don't think there has been a time where either of us have felt the need to look up a beekeeping issue online. I have watched a few beekeeping introduction and tutorial videos online out of boredom and curiosity and I have never seen one that was as helpful as having read a book. When I have to repair something on my car I don't go to the internet anymore, I've invested in service manuals for my cars because the relevant information within them is easier to find and more accurate than what a search engine would give me. Why would I endanger my life by doing something to my car that someone, or something with unknown credentials on the internet says is the right thing to do?
Books are written by human experts, not artificial intelligences; their goal in writing is to help you become an expert as well, not sell adspace. And of course books are much less convenient than the internet, you can't carry your entire library in your pocket wherever you go and more importantly you can't effectively get information from a book that you have not read which requires an up-front time commitment. If you are picking up a new skill, learning as you go through the internet, when an unforeseen problem comes up you'll be frustrated as you spend way too long searching the internet for the solution to your problem, and you'd consider yourself lucky if you even knew what to type into Google to get you on the right path to your solution. If instead you had read a book about your new skill, when you inevitably encounter a new problem you would at least have an idea of what the solution to the problem could be because you had already read about it and you would know exactly where in your book to look for more information and what kind of things to Google for more information. The up-front time cost of reading a book to help you become an expert vastly outweighs the frustration to be had while amateurishly scouring the internet for clues that you might not even have the patience to properly apply.
I've been collecting books off and on for years now and in the past few months I've found that I've shifted from mainly buying books that that I think would be entertaining to looking at books I think would be useful. I recently bought a book on emergency care and rescue, a few weeks ago I almost walked out of the store with a book about food preservation methods, survival books are always fun to look through, Polyface Designs is a book I'll eventually pick up, and books on metalworking, carpentry and gardening are also ones that I should add to my collection. What books should you buy and read? Definitely ones related to your job, then ones related to your hobbies as well. Beyond that get books on skills you'd like to develop or may just find useful to have in the future, think of the skills that matter and give yourself the resources to properly develop them, starting with a reliable source of knowledge. And remember when doing so you don't have to get the most serious or fancy books, books written for beginners, like the for dummies style of book, are perfectly fine because odds are you are a beginner at most things and a book written for you will be much better than a book written for someone who already has background knowledge.
Now of course books can't offer you everything that the internet can, I've already mentioned that you can't have your entire library in your pocket so when you are in the store trying to figure out what kind of nails to buy your books won't help you. Books probably wouldn't do too well in the Ford Lightning test I made up above either. And most notably, books also can't connect you to other ideas as well as the internet can. Just while writing this post I've given you links to several other sites filled with interesting thoughts and ideas that may give you something that I never could. The internet is a great place for discovering new ideas when those ideas are allowed to flourish, but the internet is shifting away from that because it is so easy for corporations to profit off of AI generated content. And a focus on trying to discover new things on the internet can shift to a focus on discovering new people on the internet which we shouldn't be doing, internet people are not healthy.
Now of course books aren't going to magically give you new skills you still have to practice and experience the things books teach in order to gain any new skill, but the same can be said of internet content. You can read a book on welding or you can watch a video about it either way you won't be any good at it until you've had hours of practice under your belt. Then once you've reached some level of expertise at whatever you've been learning you can put together an artisinal webpage to combat the AI dominated future our internet is doomed to have.
Can we save the internet from AI? I don't know. Is it even worth saving? I'm not sure. Would we be better off if we simply turned our backs to it and embraced books as our primary method for storing, sharing, and learning information? Yes. We should use the internet to support our learning and growth, not as its primary source.