Roger Marshall is a computer scientist, a newly minted Luddite and a cynic.
If you are an avid follower of all that has been going on in the IT world, I am sure you did not miss the news about Google planning to charge users for granting them access to the company’s AI-enabled web searches. There is an underlying presumption on the part of Google that AI-enhanced search algorithms can provide better results than their vanilla search algorithms and that users would be happy to fork over significant sums of money to that end. Much like beauty, it is all in the eyes of the beholder. X has also been operating along the same principles – charging the gullible twitterati for an ‘enhanced experience’ on Elon Musk’s social media channel. Unlike Google, the algorithms employed by X do not need to be supplemented by AI since such artificial intelligence is already supplied by the subscriber and, better yet, this intelligence is constantly evolving. Generative AI at its very best.
As has been reported in the press, both Open AI (the creator of ChatGPT) and Google have run out of publicly available data such as to be found in websites such as Wikipedia and assorted governmental and non-governmental agencies. Wikipedia data, in particular, submitted by a worldwide community of users on a purely voluntary basis, is extensively curated and, if necessary, periodically updated. Since Open AI and Google are unwilling to pay users for their data, they have now resorted to using the ‘new’ data created by their AI programmes to generate even more new data, which can then be used to create even more data, ad infinitum. This is also a version of generative AI, also at its best. As I see it, money and the reliability of the new data are the driving forces in today’s internet, both of which seem to be in short supply. But not greed.
Consider what would happen if a valid segment of information is corrupted by even a small piece of fake information or misinformation, which the AI programme fails to recognise and rectify. AI programmes and their creators have yet to explain how a particular result was arrived at. Over time, erroneous conclusions based on data contributed by ignorant or malicious users and those generated by AI learning algorithms through probabilistic and/or inferencing mechanisms will simply result in chaos in the internet information space. Isn’t this how gossip and rumour-mongering operate, to wreak lives and weaken societies? Isn’t this how a single cancerous cell, if allowed to metastasize, has the capability to kill a person rather quickly?
When one comes across any news item on the internet, the authenticity and veracity of the news item rests on who or what the primary source is. If the primary source is unavailable or inaccessible, the news item should be regarded as mere hearsay. From this perspective, it is unfathomable why Google would choose to block links from local news outlets appearing in search results other than to prevent them from charging Google usage fees for their journalism. Isn’t this a form of censorship and blackmail rolled into one? Meta, aka Facebook, in response to legislation in Canada requiring social media companies to pay print news media for content, cut off all news links to Canada. It was merely aping Google’s playbook. Meta is also heavily involved in incorporating AI into its products for ‘enhanced’ user experiences.
Given Google’s plan to charge customers for AI-enhanced search options, it is not inconceivable that in the future, the company may decide to charge users for using its search engine. The airline industry offers a good example. These days, when you buy an airline ticket, you have to pay extra for checked luggage, meals, and choice of seats – items that once used to be free. When faced with severe financial pressures brought on by government legislation, Meta, like X, may decide overnight to start charging subscribers. Since users across the world have become accustomed and addicted to Google and Google has captured 89% of the search engine market, the company is in no danger of losing too many users. The same reasoning can also be extended to Meta.
Internet users lost the privacy war some time ago. They are now on the verge of losing the data war.
In retrospect, the comment made some 50 years ago by an IBM-Australia vice president, viz., “Whoever controls information will control the world” doesn’t seem all that farfetched, does it?