The AI backlash has already begun
We're nearly a year into the supposed AI revolution and things are only moving as fast as the smoke and mirrors protecting its interests
One thing we’ve seen regularly over history is that people will, in the main, refuse to believe something can happen if it’s not in their best interests for it to happen.
Disaster capitalism has made people very rich in very different ways; if you’ve watched the Big Short, humble Cornwall Capital’s entire model was to seek out opportunities in exactly this way. The sub-prime mortgage crisis was their most high-profile success, but factoring in the likelihood of something happening that no-one really wanted to happen was their full thing.
There’s now so many gigantic organisations invested in the success of AI and its proliferation that it is now impossible to weed out the tech bros and their newsletters proclaiming that the world has changed forever, from the legitimate analysts who remain utterly unconvinced by the processes.
But AI is really amazing, right? Right??
Just last week Amazon announced its $4 Bn investment into Athropic in their quest for ‘generative AI’. But the elevator pitch to future customers is as vague as pretty much anything else you’ll hear. They say:
“Customers of all sizes and industries are using Claude on Amazon Bedrock to reimagine user experiences, reinvent their businesses, and accelerate their generative AI journeys.”
Sounds great. Except that maybe I’m stupid, but I’m not really sure what any of that means, and I don’t think many other people do either.
Jensen Heung’s Nvidia presentation to the board last week - one of the forefront AI companies - is a masterclass is speaking for two hours without saying anything at all. Meanwhile Open AI, probably the most well-known given their successful push of Chat GPT, saw their CTO, Mira Murati, fumble through the most awkward of interviews where she avoided saying their text-to-video tool, Sona, had been trained using YouTube - which would be very illegal.
And at the centre of all is Sam Altman, who is about as convincing as Elon Musk in all the wrongest of ways. His plans for AI align with quite a lot of what people who lack any credible creative thoughts want AI to be - a leveller for those bereft of original ideas. I mean, this is the type of tech guru quote that idiots think is profound.
But it’s already changing everything? Isn’t it?
Well, no, it isn’t. Chat GPT has set itself some ambitious revenue targets for 2024 for the paid version of its product but it pales into insignificance compared to the amount being invested to improve the quality of the outputs. And there’s significant competition that has opened up in the space, stealing market share.
And much of the output from Chat GPT is simply reassembled content from other pages of the internet, packaged nicely, because that’s literally what it is. It’s not intuitive, and it’s not learning, and in fact it’s getting factually worse.
Ed Zitron’s newsletter on AI comes strictly from the camp of skeptic but is highly recommended. His research shows that tools like Chat GPT have maxed out what they can learn from the internet at large, and, ironically, need original reporting - ie, journalism - in order to improve it.
But what it’s being given instead is a proliferation of nonsense, spat out on the web by non-creatives looking to overcompensate for their lack of talent.
OK, I’m a publisher, help me out here.
Ok, and now we get to Google. Google’s relationship with AI is arguably more complex than others. The likes of Altman need to continue to chip away at the grift, to prove that AI is something we all need and it needs to happen tomorrow.
Google’s ad business via its search engine is immensely profitable, but it cannot be seen to be ignoring what the tech community at least believe is progress.
On the one hand they have been flirting with Search Generative Experience (SGE), in which a form of AI chat replaces links for certain search terms. But Google has also realised that:
The generated results are grossly inadequate
The cost of running SGE is enormous
They are actively losing ad revenue because of it
AI-produced content is obliterating its search in all of the predictable ways.
So what’s next?
There’s murmurs in the SEO community that Google might have to charge users for SGE, which would be a gigantic sea change from how they have monetised any other product in their history, and it’s something that should be an overwhelmingly positive signal for publishers.
Furthermore, Google’s October core update and the ongoing March update has unquestionably caught some good sites in the crossfire, ones that likely didn’t deserve it, but it was done with the intention of crushing AI-bred garbage. While they haven’t got it right, there’s an acknowledgement that that type of content isn’t the answer to any of life’s queries.
And finally, there’s already AI pushback not only in the journalism industry, but beyond. AI was supposed to take care of the mundane tasks to allow us to be more creative. Instead the opposite is happening, and people are already sick of it.
So from the depths of last year there are shoots of positivity. The robots might not be taking over just yet - if it ever really happens at all.
Intriguing! Hopefully we will see more innovations in artificial intelligence https://servreality.com/machine-learning/ in the near future.