ChatGPT: What Could Possibly Go Wrong?
- May 22, 2023
- Andrew Keen
First as tragedy, then as farce. That’s the history of Silicon Valley.
The future is back in business here in Silicon Valley. For the last few years, all the buzz has been about blockchain technology and cryptocurrency — the ‘Web’ revolution that its supporters promised would return the internet to its original decentralised state. Today, however, the quaint nostalgia of Web3 has been swept away by the violently futuristic promise of artificial intelligence. This current revolution is being pioneered by ChatGPT, a generative AI product developed by OpenAI, a multi-billion dollar start-up backed by Microsoft and run by Sam Altman, the newly-crowned prince of Silicon Valley.
In the Philosophical Investigations, posthumously published in 1953, Ludwig Wittgenstein remarked, ‘Philosophical problems arise when language goes on holiday.’ With ChatGPT and generative AI language has, quite literally, gone on holiday. Perhaps forever. OpenAI has built an algorithm able to mimic human language with such uncanny accuracy that Microsoft co-founder, Bill Gates, believes ChatGPT is as revolutionary as mobile phones or the internet. ‘The Age of AI has begun,’ Gates pronounced about a revolutionary era in which he believes we humans are about to outsource our most unique invention, language, to machines.
Promises, promises. We’ve heard it all before. Language has been on perpetual holiday in Silicon Valley since most of us can remember. Over the last 50 years, we’ve been promised all sorts of revolutions — some vaguely real, others absurdly fantastic — by futurists like Gates. So what can we learn from the history of Silicon Valley ‘new ages’ about today’s ChatGPT generative AI revolution?
The first is that language matters. Wittgenstein was right. As Orwell reminds us in his 1946 essay ‘Politics and the English Language,’ its abuse is both a political problem and a reflection of a problematic politics.
Today, language matters more than ever. This current generative AI revolution attempts to replace — or, at least, reproduce — human language in machines. The classically-Silicon Valley utopian promise is to take the politics out of language. But language about language is unavoidably political; doubly political, in fact. And the words coming out of the mouths of revolutionaries like Sam Altman aren’t created by machines. They are human words. And, like all human language, these words have political meaning and symbolism.
The political promise of the digital revolution has always been sold by tech futurists like Steve Jobs, Tim Berners-Lee, Mark Zuckerberg, and Elon Musk in the language of justice. From Apple’s personal computers to the internet to the Google search engine and the Facebook social network, digital technology has been presented as an equaliser, a democratiser, as something fairer and more ‘open’ than before. And the language of justice is now being used by today’s generative AI revolutionaries like Altman who, not uncoincidentally, runs a company called ‘OpenAI.’
But the truth about the digital revolution of the last 50 years is often the reverse of the official Silicon Valley line. In the Valley, two plus two usually equals five. So for all the talk of more democracy, more access, more egalitarianism, the reality is a winner-take-all economy run by multi-billionaire tech entrepreneurs in which we, the people, pay with our privacy for ‘free’ services like Google and Facebook. Welcome to our tragic age of surveillance capitalism in which personal data is the new oil, and both private corporations and governments know everything about us.
First as tragedy, then as farce. That’s the history of Silicon Valley, particularly this latest wave of generative language tech companies. The whole thing could have been concocted by a tech satirist like The Circle author Dave Eggers. OpenAI is anything but open. Founded in 2015 by wealthy and powerful technologists with some vague promise to build ‘good AI,’ the once non-profit took a $10 billion dollar investment from Microsoft earlier this year, thereby valuing the now for-profit start-up at $29 billion. The intent is to include ChatGPT and other OpenAI products like its Dalle-E digital image algorithm in closed Microsoft products like its Bing search engine and its Cloud services.
Microsoft’s goal is to displace Google as the most lucrative internet software company, in the same way Google replaced Microsoft 20 years ago. Meanwhile, as a reaction to the ChatGPT hullabaloo, Google is pouring tens of billions of dollars into Bard, its own Generative AI product. So the likely consequence of today’s AI arms race is an even more top-heavy winner-take-all economy in which one or two trillion dollar tech firms sell ‘empowering’ and ‘democratising’ technology to you and me. First as Web 2.0 tragedy, then as Generative AI farce.
But there is one, potentially significant, difference between the origins of the Web 2.0 revolution in the early 2000s and today’s Generative AI revolution. Back then, there was very little responsible conversation about the dangerous consequences of transforming everyone into an internet publisher. So all we heard were Orwellian promises about the empowering ‘Long Tail’ of content and other self-serving garbage. Today, however, there are more responsible voices — like the so-called ‘Godfather’ of AI, the former Google tech exec Geoffrey Hinton — warning about the unintended consequences of our AI age.
Even Altman is acknowledging the dangers of his revolution. In Congressional testimony this week in Washington DC, he confessed that new technologies ‘will destroy jobs.’ In the Age of AI we will be on holiday all time.
But who will pay for this brave new AI world of worklessness? Not surprisingly, Altman is also the founder of WorldCoin, a creepy start-up creating a global identification system through iris scans to give out a ‘free’ global currency. WorldCoin is artificial intelligence meets universal basic income. What could possibly go wrong?