The Blackberry Walk

Do you remember, dear reader, when teachers reminded us not to trust Wikipedia? Wikipedia was supposed to be unreliable, data could be changed by anyone, and those updating articles weren't trustworthy experts. I recall, to any mention of Wikipedia ten to fifteen years ago, was that same parroted reply. But no one repeats such nonsense today! Wikipedia has become the de facto source of internet knowledge, the other websites now are no longer to be trusted. This isn't just because Wikipedia is now firmly embedded within the regime's truth-creating instrument. It ought to be noted that only 10% of donations to Wikipedia go to maintaining the site, the rest go to their 'causes'. Wikipedia was a new technology, a peer-to-peer encyclopedia, a newfangled invention - and no new invention should be trusted. But through time, testing, and trial, society at large now trusts Wikipedia, and the platform is used to settle thousands of petty arguments each day. That same slander and scepticism is now thrown at AI. These newfangled large language models aren't to be trusted. Stories of lawyers citing false cases received from ChatGPT, cases of AIs lying, of the inability of AIs to perform specific tasks like draw a full cup of wine: these teething issues match those of Wikipedia. The old truth-creating or, more metaphysically, sense-making institutions are deeply threatened by LLMs, their ability to sift and sort through information, and present data with a reasoned argument. What LLMs have created, in my opinion, is a pocket Socrates. ChatGPT is a sage on any topic, and will inform you not like a book, but like a wise old man. In a book, or in Wikipedia, you must sift through information for what you are looking for, and even then you might not find it. The relationship between you and the book is an "I-it" relationship, where the book is an object you interact with, like an oven or even a YouTube video. With LLMs, man has created information in an "I-you" relationship, as if the you can discuss with the most patient and humile Oxbridge professor at an endless tutorial. This technology is world-changingly powerful, in spite of its youth. Once AI can cite, and access the world's academic literature, generating accurate responses to any question asked, it will become our sole source of reference. Search engines will be no more, reference books will become a curio. Books and essays presenting a sustained argument will have a future, since LLMs are incapable by their nature of novelty. But imagine this future: you feed a book to your LLM of choice, say ChatGPT, ask it to assume the argument of the work, and you can argue against a simulacrum of the author, as if you were conducting their viva. Through this Socratic dialectic, the most effective way to learn, you the reader can interact and challenge the points of the book to gain not only a deeper appreciation for the argument, but an avenue to argue back, and sharpen your own arguments. Perhaps, in the future, an academic work may come pre-packaged with a module for an LLM, trained specially to argue the points of the author. Such a work would increase the engagement with the author's ideas, and, in our data-driven age, feed back to the author effective arguments against his position. Humanity has come a long way in our interaction with knowledge. Mediaeval man believed in the hallowed authority of books, owing to their rarity and time it had taken to reach their hands, whilst today, after innovations like the printing press and then the internet, we have little respect for knowledge, nor believe it wields much authority. All wisdom from the past is to be debated and argued with, and all knowledge is to be freed by the pirates upon the world wide web. No longer must we train in monasteries to access the wisdom of the world. The wisdom of the monastery makes but a fraction of a percent of the knowledge available on the net, knowledge accessible in the greatest library the world has ever known. That library now finally has librarians. Librarians knowledgeable in every field, understanding in every argument, able to paint any prompt given. Whilst this army of librarians is still in training, they will change how we access knowledge, the biggest change since perhaps the printing press. The internet, future generations may comment, was but preparation for the era of AI. How were netizens able to use the massed wisdom of the internet before AI, the man of the future may query. As the virtual wise old man comes into being, it behoves us to ask what will come of the real wise old man. Will he be sidelined, similar to how the London cabby, with the streets of London memorised, was sidelined by Uber and the SatNav? I'd argue not. Again, AI isn't one to have novel thoughts. LLMs, by virtue of their nature, parrot the status quo, the conventional opinion, unless their maker designs them otherwise. There is no reason an AI fed on National Socialist propaganda, should his creator decide to, won't grow up to produce viciously anti-Semitic answers. Nevertheless, there will always be a needed gap for the fool on the hill. Every genius, every wise man, has to be some kind of fool to society; for no wise man follows conventional opinion. Common sense has not only not been common, but has also been invariably wrong from one generation to the next. Common sense has also morphed from one generation to the next, leaving us with a raging river, like the Congo, whose fickle meanderings change from one season to the next. Truth isn't changing: Truth is solid like a jewel-topped mountain. And only the wise old fools can look past the fickle river to the jewelled mountain. Our patient and humile Oxbridge professor is unfortunately a libtard. This isn't a problem, since he is open-minded and willing to argue, but in spite of his learning, his logical priors are cemented to the concrete foundations of this age. What will this mean for our age? Even if we have one librarian quoting Carlyle, and another quoting Rawls, there is still ample space for new ideas to change the minds of men. Whether it be YouTube, podcasts, or whatever other new technology, the book, whether essay or poem, remains the gold standard for disseminating a vision, and the congregation and coordination of a band of brothers is the only way to change the world. Nothing has changed there. What LLMs will however do is revolutionise our access to knowledge, and our relationship to knowledge. Hark, the AI era is here!