Nothings Inevitable

Things that are inevitable rarely announce themselves as inevitable. In fact, AI seems to already be slowly dying and here's why you should help it go away faster.

Over the past few weeks I've sadly seen more and more people, especially artists wheel out the same line, AI's here to stay so why shouldn't I use it? AI's inevitable so maybe we should all learn how to live with it. AI's just a tool in my toolbelt. Grumbling to themselves as they drain a nearby lake to generate some slop for someone to scroll past.

Despite all the yelling by billionaires, slop boosters and bosses alike, I've never used genAI and I don't plan to. Somehow fours years since ChatGPT burst into the public conscious I haven't been left behind or lost my job a chatbot. And I think I'm still pretty good at telling when someone has.

This isn't a brag, every other ethical concern aside, it basically boils down to the fact that I don't trust it and I hate its outputs. The tiny bits of work it might be useful for, are actually things I enjoy doing like reading and researching, analysing data or responding to people who have taken the time to reach out. While I've not been shy about my complete disdain for AI, I can empathise with those who seem to think its march to overrun every facet of our lives is inevitable.

It's all we've heard since the launch of OpenAI's ChatGPT on November 30, 2022. ChatGPT the other 'genAI' tools that arrived around the same time in 2022 were and are interesting pieces of software but they have in large failed to deliver on every metric. Every half baked white paper perpetuated by OpenAI or Anthropic has been exposed to be a complete fallacy as time caught up to their timelines again and again. From the destruction of Moore's Law to AGI and now Recursive self-improvement demonstrating the moving goal posts and increasing abstraction of what can only be referred to as an attempt by Silicone Valley to extract every drop of profit out of society before collapsing in on its self.

Nowhere is this complete lack of delivery from the slop bots more apparent then in their own financials, as per Ed Zitron's reporting.

"OpenAI had $13.07 billion in revenue and $34 billion in costs. $867 million of its revenue came from SoftBank, and $303 million came from Microsoft." Ed Zitron.

OpenAI after almost four years in the slop game is still losing almost $20 billion a year. The thing we've been told is inevitable, that requires free access to the labor of millions of artists, trillions in dollars and an endless supply of water and power is still losing $20 billion a year. Not even the biggest, wealthiest corporations and individuals with a license to burn money and free access to resources can make this software profitable.

A time will come when the money runs out, unfortunately that will come before a single government official makes a genuine move to protect the unpaid labor of artists or our collective natural resources. In fact Australia's current government is considering the opposite. As this happens, the venture capital, petro-states and billionaires will try overinflate some other trend or try short each other into oblivion in the style of the 2007 Global Financial Crisis (GFC). I don't know when the 'AI bubble' will pop, no one does, irrational markets to tend act irrational for longer than rational ones but it will pop and it will probably take your super fund with it.

Speaking of the 2007 GFC, you know what else happened in 2007? Steve Jobs announced the first iPhone, on a stage in San Franciso on January 7th to be precise. If you watch the Steve Job's announcement of the iPhone, the thing that really revolutionised it as a product was getting rid of the keyboard and developing the first decent touch screen. It was the third product in a decades long run by Apple that truely disrupted multiple industries from computing to music to the phone. They also benefited from the aforementioned 2007 GFC which was going to kick off only a couple months later. As stocks across almost every other industry tumbled Silicone Valley largely held and boomed, this wasn't just due to Apple of course but a swath of tech companies poised to capitilise off of the GFC.

I'm speed running and omitting a lot, but Silicon Valley kind of lulled through the 2010's but surely enough Google, Meta and kind of new comer Tesla boomed. Then came 2020 COVID pandemic and basically every tech company in the world exploded and bloated. But this is where we can start to see something new emerge from Silicone Valley. They'd run out of hypergrowth ideas. Since 2020 the future has come and gone multiple times in the forms of various pyramid schemes from crypto and NFTS to the metaverse and now of course genAI.

Zuckerberg famously renamed Facebook to Meta during this time, having just recently shelved his metaverse project after burning somewhere around $70billion on the failed project. Can you imagine what you could do if you were allowed to spend $70billion on whatever you wanted too?

Ford has just rehired 300 engineers after AI failed to deliver. Uber, famous burner of cash, has already burnt through its AI budget for 2026 with minimal to no productivity increase. All of this to say that despite everything being screamed at you from trillion dollar corporations or your manager, AI isn't the next iPhone or Facebook, it isn't inevitable and might actually go away. Soon too.

Large swaths of the broader community are beginning to realise the consequences of the mass rollout of new data centres across the world. Until very recently many in so called Australia assumed this was largely a problem for elsewhere but are now noticing these giant humming buildings taking up prime space in their cities and communities.

Australia's number one builder of data centres, with a healthy dose of greenwashing, is NextDC. A few weeks back I passed one of their recent build outs on the way to see a band play at The Tivoli. It sat in the middle of Fortitude Valley at nine stories high and taking up an entire city block. This same company, NextDC is currently in the early stages of construction of another 'AI datacentre' just a few streets over from my own house.

Something that has finally been brought to the attention of a the broader community thanks to a few kids posting up in the city centre to protest it, albeit about three years too late to stop its inner city construction. Now we need data centres, there are already more than 160 facilities scattered across the country enabling a lot of amazing tech. But there are currently 100 further data centres either planned or in development all largely focussed on 'AI'.

Datacentres built specifically for AI can't really be efficiently used for anything else but AI compute, so what happens to these buildings if the demand for genAI crashes? Buildings that are designed to require the least amount of people to operate them as possible. Giant white elephant buildings in our cities and communities that don't activate those areas or contribute, purely existing to extract resources.

But don't lose hope. According to Data Centre Watch, in the USA during only quarter one of 2026 at least 75 data centre projects worth approximately $130 billion were disrupted by local opposition.

"The back half of twenty twenty five marked a turning point, as data center opposition emerged as a national level narrative that showed the AI industry can no longer see the fights as individual zoning disputes. It is now reshaping elections, regulation, and site viability nationwide."

What's more interesting about the opposition to data centres in communities , is that its bringing people from across political divides to oppose local construction projects. Uniting disparate and conflicting communities not only around opposition to data centre build out but also to imagine what they want for their communities.

How long do we happily sit around and hear that there isn't enough money for things like healthcare, education, libraries or the arts when all levels of government actively subsidise or fund the building of infrastructure for trillion dollar companies like OpenAI and Anthropic.

Data centres are popping up all over so called Australia, in the middle of our communities and cities in the middle of housing crisis no less. But there's been backlash and its building. For instance, Anthony Albanese's alleged $50billion AI deal aims to see Australia's copyright laws watered down to try get AI companies to commit $50 billion to data centres in Australia, with a measly and further $350 million each year into a fund for Australian artists to mine Australian creative content.

Albo's AI deal is kind of the equivalent of trying to convince the makers of the Titanic to build more Titanic's using slaves as the Titanic hits the iceberg. So long as they contribute some money to slave fund of course. Remember how I began this rant by mentioning that OpenAI is still losing almost $20 billion a year? Meanwhile Microsoft's capital expenditure spending for 2026 is 23% higher than expected as they lay off 4,800 jobs to reign in spending.

Who exactly is going to pay for the $50billion in data centres Albo? Will that $350million ever eventuate when you look at those numbers? What if, and I hope it does, community sentiment around data centre builds follows the same trend in Australia as in the US and communities across Australia stop opposing and stopping them being built?

If your upset at Albo and Aus Labor making it easier for foreign companies to profit off of creative labour, then APRA AMCOS have an open letter you can sign against the watering down of Australia's copyright laws. The Greens are calling on a moratorium on the building and approval of new data centres in Australia. Those kids have managed to make a few aware of giant-resource extracting building being built in the community.

I've never been a big fan of blaming the consumer, especially when much of the products we're forced to use have eroded the place of other often more sustainable and effective products purely because the profit margin was better. Swapping plastic straws for cardboard ones did nothing to reduce the microplastics in our oceans but it did enable multi-national corporations to cut costs and greenwash.

Despite the aforementioned yelling about being left behind, and some very early play with Dalle, I haven't used genAI once. I've never used ChatGPT. I've never used it to write an email or soundboard an idea and I have no desire too. It's okay if you have, I really don't care if you have or do, the point is more it hasn't left me behind in some race to ultimate productivity.

GenAI and social media platforms are different though. These trillion dollar companies rely not only on a steady flow of users, spending more and more time but also an endless supply of new users. If public sentiment turns enough they crumble. Pour one out for MySpace. Reducing or cutting off your usage, especially in a public way can have real impact.

Just last week Meta has announced a new feature on Insta that allows other users to generate AI content based on photos you've posted while it also being revealed that their data centre potentially also poisoned a city's water supply with Legionella Bacteria. Both have received widespread opposition and condemnation. So much so Meta rolled back their most recent creepy AI feature by the end of the week. I just can't wait until Zuck launches the AI version of himself that he's going to force Meta employees to talk to.

Other dumb AI things that have died; Sora, Atlas, "adult mode" by ChatGPT, and most tellingly OpenAI IPO has been indefinitely delayed because its still losing $20billion a year and Musk's SpaceX took all the money and is slowly tanking.

Just as Meta's creepy AI products aren't inevitable. AI isn't inevitable, a lot of it is already starting to die and there's a bunch you can do to get involved to chip away at the ground its stolen. Just make sure you do it loudly.

Inspired by Matt Chun and Sophia Cai, I recently added a declaration of not using AI to my email signature. It's a small flag planted on the side of I've read your email and I'm going to take five minutes and genuinely reply to it. There's an example below that you can also use if you'd like as well.

Here's to being left behind.

P.S If you use genAI for your event poster, venue or anything visible and noticeable I won't go.

Email Signature Example From Sophia Cai: Human written: I do not use generative artificial intelligence (genAI) to research, develop, edit or refine my work. No genAI was used to write or edit this email. As an artist and arts worker I do not consent to my work being used for the purposes of training AI technologies.