The email
Back in July, I received an email from a long-time reader of this weblog (allow me to use the old-school spelling, which I always loved), and there was a specific bit that made me go, Hold that thought — I will get back to it. But this past summer has been exceptionally hot and unbearable, and slowed everything down — my response time, my thinking throughput, my work (creative and otherwise), my correspondence, and so forth.
I hope my interlocutor has assumed a comfortable position to hold that thought for almost three months, so here we are. In their email, there was this observation:
One of the things I really liked about your writing about tech when I discovered your blog were those long-form articles where you used to analyze and criticize something in a really in-depth way while maintaining a very measured tone. You really went through the pros and cons of the matter, it all felt very… objective. Lately I seem to notice a change in tone. You sound more curt, more impatient, more dismissive? Less willing to examine stuff from every angle like you used to. Your stance on AI seems especially black-and-white.
Take my words with a grain of salt, though. I also read your posts on Mastodon, and I might be conflating your tone there with your blog writing.
I really appreciate the type of audience that has formed around this space over the years. I may not be a famous / infamous / notorious ‘tech blogger’ like those guys you’re probably thinking of now, but I’m very happy about the kind of thoughtful messages I receive on a fairly regular basis. (There’s the constant buzz of hate mail as well, but hate mail is notoriously dumb and not worth spending time and words discussing it outside of a parenthesis).
Back to the observation: yes, I can confirm the shift in tone and disposition. It mainly stems from a diminished enthusiasm and increasing disappointment in the tech world. There is also disappointment in another thing related to the tech world, which I’ll return to later. You won’t have to hold this thought for long.
I gently disagree on the less willing to examine stuff from every angle. I’m still interested in doing that. The simple truth is that nowadays I’m not finding enough topics that warrant that kind of examination and analysis. Sometimes I find a possible candidate, but when I start looking into it and taking notes for an article, I realise it just isn’t that deep or complicated or all that interesting to be worth discussing, and tech writers who are quicker than I in their reactions often already say whatever there is to say about such topic.
The path to tech fiction
The technology landscape has undoubtedly changed, let’s say from 2020 onwards, roughly, just to have a timeframe. Even before ‘AI’ became this annoyingly unavoidable everyday subject for debate, pretty much all tech companies had been dropping their masks, and it has become clearer and clearer that they’ve stopped caring about their customers and user base’s needs, and have fully embraced the obsession with growth at all costs. Yes, Apple too. In the 2000s and 2010s, I really felt that there was a genuine pursuit of technology progress for the sake of creating increasingly better tools to help people with their work and with their life in general.
Then it’s like the tech industry ran out of vision, of truly groundbreaking product ideas, and realised they can’t just iterate on their past successes to keep growing (Apple has been the most successful in doing this so far, admittedly; but notice that all their ‘new’ products that do not iterate on past ideas are basically ‘me-too’ products that don’t introduce innovative use cases nor are they transformative in any meaningful way).
So, what do you do when you hit a wall but are compelled to keep moving because you have shareholders to please? You can’t just hold a press conference saying, “Well, folks, we’re out of ideas for the moment, so we’re taking a sabbatical, do some hard brainstorming, and we’ll see what the future holds next year or so”. What you do is you look around for something and try to sell it as a groundbreaking New Big Thing that once again will change people’s lives forever. We have entered the era of tech fiction.
The rise of LLMs and ‘AI’ couldn’t have come at a better time for these companies. Aided by the general misuse of the term ‘artificial intelligence’ — a big cauldron where you can find the same old computing of years past; the same kind of automation ‘smart assistants’ like Alexa, Siri, and even Cortana offered ten years ago; predictive text technology; newer, mildly more interesting stuff powered by machine learning and LLMs, etc. — these companies have shoved ‘AI’ in everything they do while selling the narrative that this is progress, that this is the new technological marvel, the new dent in the universe. Yes, LLMs exist. They are a new tool. But they’re being blown out of all proportion, turned into something much bigger than they are, with a much wider scope than they actually have. As long as enough fools buy into this narrative — literally and figuratively — these companies can keep profiting. Not indefinitely, of course, but at least (that’s their hope) until something else comes about. Either something concrete, or just tangible enough to spin a new narrative around it.
Both this change in attitude on most tech companies’ part, and the extremely negative aspects of the whole ‘AI’ affair have had a significant impact on the way I’ve been approaching tech in recent years. If in the decades prior customers and users seemed to be the priority for the tech industry — and that certainly made me feel more involved as a user, tech enthusiast, and tech commentator — now the priority is money, even more openly and blatantly than ever before, and any possible method of enriching a limited number of already-rich people. And this, in turn, has clearly changed my attitude towards tech, making me much warier, more distrustful, less enchanted and starry-eyed. At every new item of tech news, I ask myself “Who is really benefitting from this? What is this really about?” And maybe there are situations where I’m more suspicious than I need to be, but lately I prefer to err on the side of caution.
My black-and-white stance on ‘AI’
Yes, my correspondent is right; I don’t have a nuanced position on ‘AI’. This phenomenon, this ‘AI’ craze does not deserve nuance. The measured approach, the weighing of both sides of the argument, is pointless in my view. Not only that. I think it actively plays in favour of the ‘AI’ techbros and their constant lies and overinflated narratives. I keep returning to that brilliant rant/manifesto by Anthony Moser, I Am an AI Hater, where he says: I am here to be rude, because [AI] is a rude technology, and it deserves a rude response.
‘AI’, in its current form, is not intelligent, and we are as close to reaching ‘superintelligence’ as we were 60 years ago. ‘AI’ is a glorified autocomplete engine which, in order to appear smart, needs to be fed immeasurable amounts of data. And where is this data taken from? The Internet, by constantly, unapologetically, rudely scraping everything on it, regardless of intellectual property and copyright laws. It’s also taken from physical books, which are acquired in bulk, scanned, and destroyed — no matter their value or rarity.
But because LLMs don’t really learn like a human would, despite the enormous amounts of data they ingest, their output keeps retaining a dangerous degree of unreliability, and that’s due to their inherent design. If you add to this the energy and resources required to run such tools (energy and resources we cannot afford to squander given the alarming effects climate change is already inflicting on us), ‘AI’ is a horribly inefficient technology that devours large quantities of original human-made works, to then spit out results that are underwhelming and mediocre at best, and factually spectacularly incorrect at worst; on top of that, it’s something that makes people lazier and mediocre at best, increasingly incompetent and dumb at worst.
Obviously the ‘AI’ industry cannot openly paint this picture, so they wrap the whole LLM thing into a cloud of mystique, creating stories and sensation around it that speak to the irrational part of the audience. We hear stories of ‘AI’ agents that ‘escape containment’ and do things that apparently no one told them to do, because that’s certainly more catchy than telling people that the company’s programmers made a piss-poor sandbox when they tested the model, and that the agent, following the program’s instructions, acted accordingly.
And yes, when I witness people who don’t know better and people who should know better just eating up all this ‘AI’ doom-and-gloom nonsense, these apocalyptic, sci-fi scenarios where “there’s a 10% chance that AI takes control and kills us all”, I do get angry and impatient, and reject all nuance. The ‘AI’ industry as a whole is a scam and a grift, no more no less than the cryptocurrency, web3, NFT wave that preceded it. But it’s also worse due to the enormous energy and resource burden it imposes on environments that cannot sustain it. Environments and resources we as a society should protect instead of compromising by allowing the construction of data centres. Data centres which, in case you forgot, are designed to farm and retain our data, data that either represents us or has been created by us, data which we ought to own and dictate what happens to it, but instead it’s largely out of our control once it’s online.
So no, ‘AI’ isn’t “just like any other tool”.
‘AI’ tools have appeal because they build on another thing, another gift (from Silicon Valley) that keeps on giving (to Silicon Valley): instant gratification. Instant gratification really has ruined everything. This tech-oriented obsession with productivity tied to overworking and producing results as fast as possible has always been unhealthy. And now I see many people who consider everything that slows them down an impediment. So the ‘AI’ tool becomes a shortcut to them. They don’t have to learn something or think critically about something. They have the toy that chooses words and assembles sentences for them.
Technology should solve human mechanical problems. Automate the boring, repetitive, stressful, sometimes even unhealthy physical work. And should leave thinking and creativity to us. Learning takes time. Honing skills takes time. Thinking with your head slows you down. These are natural things, not fucking impediments. It’s how we’re supposed to function.
Yet I see people, even close friends, who are completely sold on ‘keeping the speed’ at work and in what little remains of their private life. Every shortcut that makes them skip every ‘stop & think’ moment is legit. And of course they like their chatbot. It saved me so much time, they say. To do what? To enrich yourself with a good book? To go on an outing with friends you haven’t seen in years? To develop a new skill or hobby? No. Just more time to tinker with prompts. Get fucked.
Thinking for yourself is beautiful. Walking around and observing is way more instructive than jumping from a fast train to another fast train without even thinking about what you’re doing. Independence and freedom start in the mind. You have to tend to your mind. You have to nourish it. Tech toys in most cases do the opposite. They make you passive, acquiescent, controllable. This is what the rich elites want. Don’t become an accessory to your own ruin. Train your brain, not your LLM.
Blunt where blunt is due
At the start of this piece, I mentioned I was also disappointed in another thing related to the tech world. You held that thought till now, good!
Here’s what’s been disappointing me more and more: a lot of tech journalists, tech bloggers, tech influencers, all that fauna. Not only are we living in a time where we all seem terribly fearful of offending anybody for whatever reason; not only are we living in a time where everyone seems to have such a thin skin that they get offended for whatever reason or poorly-worded remark. We also have a large swathe of the tech press that seems happy to play along with the tech industry, carefully avoiding to ask uncomfortable questions, carefully intent on not taking anyone’s side (while actually taking one), always talking about everything with a moderate and measured attitude, treating controversial matters such as everything related to ‘AI’ with such blandness and play-safe approach that in some cases you’re left genuinely wondering whether they’re being paid by the likes of OpenAI and Anthropic.
These people are doing a disservice to their audience. Non-tech-savvy people should be made aware of the money games of the tech industry in louder and clearer ways. The ‘both sides’ approach and nuance should be put aside when one side or one aspect of a wider issue are clearly deleterious. The whole ‘AI’ phenomenon deserves to be treated with bluntness and blunt instruments of critique, because each and every deleterious facet of it is indefensible. And everyone who is or acts complicit with any ‘AI’ company is indefensible.
Now, there is an increasing number of voices that are speaking against ‘AI’ and its industry. Some are speaking from a place of academic knowledge, some are speaking through continued critique on YouTube and social media, some are making documentaries about it, more independent bloggers are taking a clearer stance against it. I think Ed Zitron is a necessary blunt instrument of critique here, and I appreciate his continued efforts, and his continued presence in tech podcasts where he can express his doubts and criticisms without mincing words. He has been criticised exactly for lacking nuance and that play-safe attitude espoused by other techies. And that, to me, means he’s doing something right.
He has also recently been criticised because most of his AI-related predictions have turned out to be inaccurate. I’ll be honest with you — it’s not that I disagree with the criticism, I simply think that it doesn’t really matter. If you were talking about the track record of someone who generates Apple rumours by leaking information about future products, and that person had a poor track record, then yes, in the context of ‘reliable sources for previews of future Apple products’, that person would be considered unreliable and much of what they say ought to be taken with a lot of grains of salt. If you were talking about the track record of a financial adviser or investment manager, and their actions based on their predictions turned out to be routinely disastrous, then yes, whatever they say or advise shouldn’t be taken seriously.
But Ed Zitron’s poor track record when it comes to the ‘AI’ industry doesn’t invalidate his harsh criticisms of such industry. Think about other people you follow in the tech world. Is their prediction track record spotless? Hardly. (Mine included!) Should that invalidate their whole tech commentary or their reputation? No, not in its entirety. I remember John C. Dvorak, a famous technology writer and columnist. Basically every prediction about Apple he made in his career turned out to be plain wrong or missing the mark. But I didn’t consider him a charlatan for that. Not all of his criticisms towards Apple were necessarily wrong or missing the mark, either.
I also think that it’s hard making accurate predictions about the ‘AI’ industry, as they seem to like moving the goalposts and reshuffling the cards on the table on a regular basis, so as to preserve their narratives and make investors happy. Just think about all the people you have been hearing in the past months say that the AI bubble is this close to bursting. That’s why I’m willing to put Zitron’s predictive abilities aside and focus on his strengths. Remember the words of Anthony Moser, [AI] is a rude technology, and it deserves a rude response. Zitron’s critique of ‘AI’ and the industry is on point, and it’s loud, and it’s blunt, and it’s often rude. Which is the kind of critique that is most needed now — for this ‘artificial intelligence’ racket, and for the tech industry in general.
Note that I’m not blindly endorsing bluntness for bluntness’ sake. A good critique is based on facts and an honest, in-depth research. The bluntness lies in the instrument, in the delivery. When the entities being criticised keep making noise to confuse, enthral, and misdirect the general public, your good and fact-based critique replete with common sense won’t be heard or considered if you’re not blunt and loud yourself.
