Silicon Valley produces "innovation" without considering how they practically solve problems. AI is the latest but certainly not the last example.
I wrote recently about the false innovation Silicon Valley repeatedly peddles which inevitably turns out to be far less revolutionary than advertised as it becomes clear that the “innovators” know far more about tech than about business. For a currently relevant example, consider Artificial Intelligence (AI). As regular readers will know, I’m an ardent AI sceptic, not just because I think the method to train AI is essentially hyper-industrial intellectual property theft targeting both media conglomerates and regular people’s social media posts. I’m also a sceptic because it seems like a solution in need of a problem.
The term AI is fuzzy, it originates in science fiction where it means a sentient and sapient being created by humans, most commonly a robot or some form of other computerised mind. It’s a great concept in sci-fi as it let’s authors play with the morality and ethics of artificial or synthetic life. In our universe however, AI is essentially a machine learning algorithm which is trained on a large set of data to generate an output based on an input. Or in plain English: ChatGPT generates a text based on your prompts, and Midjourney generates an image based on your prompt.
On first glance that does sound generally useful; indeed there have for centuries whose full time jobs were to be scribes, typists, secretaries and other similar roles all about writing text for someone else. If ChatGPT worked as well as a person performing the task of a secretary then the fears of mass layoffs would be warranted, but few people’s jobs can be automated by ChatGPT because most human writing has a clear point, a message. ChatGPT and other AI systems works through probabilities, so while it can write text it cannot write a text with intentionality.
Unlike the Ais of sci-fi real AI cannot think for itself which makes it far less useful than it appears at first glance. If you have to edit or input several prompts to get a good ~500 word text than it would have come out better if you just wrote it yourself. Of course, there are situations where a highly standardised and very bland text is useful, such as in customer support contexts. But at that point, how much time is save by having an AI write the text instead of using a template? Ais are prone to what is euphemistically called “hallucinations” where the AI just makes something up. There is no risk of your gmail template making something up.
The real nail in the coffin for AI as a useful everyday technology is the industrial amounts of processing power, and by extension energy, it uses. There have been reports recently of people like Sam Altman and organisations like Google being interested in small scale nuclear power to meet the projected power needs for the supposedly coming AI revolution. Which begs the question: is it really better to use very power intensive bots trained on stolen data to write formulaic text which has a built in, and at least as of the time of writing, seemingly unfixable quirk where it has a small but non-trivial risk of lying inexplicably than to hire people to do that work with the help of templates?
That was obviously a very leading question, but as I’ve argued plenty of times on the blog it just appears that Silicon Valley is so caught up in the technical and engineering problems they’re solving to consider the practical uses. To be fair, AI is a very technically impressive technology and a step forward for computer science as a discipline to be sure, but technical complexity is only useful when it allows us to solve important problems. And even then, sometimes a lower tech option is more commercially viable. The Concorde is an impressive plane but it wasn’t profitable to operate, and no matter how technically impressive it is it won’t change the fact that it wasn’t profitable.
I want to end by saying that I’m not against AI as a concept, just that I think the Silicon Valley hype machine is once again promising the moon and stars without considering if it can deliver. AI is an impressive technology without much practical use; such inventions belong at universities where the pursuit of knowledge is the goal instead of in the market.
If you liked this post you can read a previous post about here or the rest of my writings here. It'd mean a lot to me if you recommended the blog to a friend or coworker. Come back next Monday for a new post!
I've always been interested in politics, economics, and the interplay between. The blog is a place for me to explore different ideas and concepts relating to economics or politics, be that national or international. The goal for the blog is to make you think; to provide new perspectives.
Written by Karl Johansson
Sources:
Cover photo by Tara Winstead from Pexels, edited by Karl Johansson