INTRODUCTION
If you’re like me, you’re a little tired of hearing about Artificial Intelligence.
As the Creative Director of a small digital agency and being plugged into the tech scene, I’ve been listening to many interviews from the Executives and CEO’s of big AI companies.
Within these interviews, they always seem quite candid about job losses, with some actively making it their goal to replace all human labour. They consistently hype up their own model’s performance, while also skirting around the risks and every other issue associated with the technology.
While on the ground, I’ve also been apart of several meetings where excited marketing managers are keen to utilise AI in every part of their strategy. In many ways, as a cost saving exercise and also to reduce the time and resources required to complete a project.
While not thinking of myself as Anti-AI, I’ve had many reservations to the shift so far:
First, while becoming increasingly convincing, I do not enjoying consuming AI-generated content. Many YouTube videos now use an AI-voiceover, and I find myself switching off quickly. One channel was even using an entirely AI-generated persona, complete with switching camera angles and different environments across videos. This one (scarily) took me a moment to recognise. But once I did, I lost my connection to the content and moved along.
Secondly, I personally haven’t found many good ongoing use-cases for AI. Because while companies keep making big promises on their products, they often fall short. Being nearly there, but not quite. And sadly, this just isn’t good enough. Plus the companies which do promise ‘agentic’ behaviour require an immense trade-off with both privacy and trust; needing to be plugged into your entire business-backend in order to create any meaningful value.
Right now, that risk is a little too high for me.
Therefore, like most people, I’ve used chatbots such as ChatGPT or Claude for various (light) purposes and testing over the last couple of years. I’ve generated a couple of images for pitch decks when I was struggling to find exactly what I required, and one time I utilised it for early concept ideation on a project.
Despite some interesting suggestions, this has not become apart of my regular process.
Lastly, AI has many environmental impacts too. While many know about the high water and electricity demands of these AI data centres, sadly it gets even worse.
There are reports of companies such as Microsoft, Amazon and Google building data centres in developing economies and utilising significant amounts of their clean drinking water to cool them.
Additionally, as AI’s usage increases, this only worsens these environmental impacts. Data centres are not able to be powered in sustainable ways due to their high energy requirements, and even Google has seen an 11 percent increase in their total carbon footprint in 2024. This is also why CEO of OpenAI, Sam Altman, is funding Nuclear Power in the United States to help reduce energy costs for his own business interests, with little regard for the larger consequences.
Even if you’re not an environmentalist like me, I see a lot of issues with these kinds of business models. One of the biggest benefits of creating any business is economies of scale. It’s the reason you can buy a smartphone in 2025, with more power than a laptop of 10 years ago, for under a thousand dollars.
Scale usually allows you to significantly reduce the cost of the supply chain.
Therefore, when you create a type of business where supply costs increase with usage, you often have a misalignment of incentives to your end customer.
Spotify is a perfect example here. The more you listen to your favourite artists, the more the company has to pay out royalties to record labels, effectively making it harder for the business to keep growing its revenue. In response, Spotify has invested in several strategies including offering musicians advertising opportunities within the app, and even commissions its own music under the name of Perfect Fit Content (PFC). This is where the company pays session musicians to produce something similar to popular tracks (often in genres such as Lo-Fi Beats or Jazz) and then inserts them into curated playlists. With Spotify owning the masters, they pay less royalties, but in return the user experience arguably suffers.
The simple fact is, at least right now, services like ChatGPT can’t charge enough for each user in order to cover its costs. This is all fine in a company’s growth stage, as we saw with Netflix, who burnt through billions establishing themselves. However, eventually investors will demand returns, and this will likely require big changes in the underlying business model.
For those creating reliance on AI for their businesses now, the enshitification of these services might be a painful process. Just like using this word in daily life… Seriously, I hate it!
Especially as the price of AI could eventually scale to cost the same, if not more, as hiring actual human labour. Of course these ‘employees’ are significantly more efficient than humans as they don’t require breaks and can work around the clock.
However, it’s still an interesting thought exercise. If your subscription to an AI agent would soon cost the same as a salaried worker, would you still choose it over a human?
ANOTHER SIDE
Last week, Monash University held an event for their alumni on ‘The Changing Face of Business’.
It included talks from speakers such as Professor Simon Angus, Professor Mariano (Pitòsh) Heyden, and Q&A with General Manager of AI for Carsales, Angus Nalwan and CMO of Google, Suzana Ristevski.
This was an important moment for me because it uncovered exactly where the source of my AI-Fatigue had been coming from:
Being constantly sold to through fear and hype.
Away from the throes of these dystopian narratives crafted by AI-startups, the conversation on the underlying technology felt exciting and energising. The focus was more on how Artificial Intelligence works and how to best utilise it. This is in comparison to the common hype-narrative that AI was coming to upend all of our lives.
Sitting in that audience, the possibilities with AI seemed endless and I actually felt privilege to be living through this transition. Mainly because being here now means we all have an opportunity to help shape the direction of the future.
It made me realise that many of the products that AI-companies are trying to sell us are merely solutions in search of a problem. However, there are real possibilities for AI to make a difference within companies and for the individuals apart of their teams.
To illustrate one particular use-case, I’ll draw on some of my own previous experience working at Apple many years ago. I can tell you Australian phone carrier Telstra had an extremely complicated back-end system for managing their consumer phone plans.
It was aging, slow, inflexible and felt like a security nightmare waiting to happen.
In fact, when the Apple Watch with Cellular was launched in 2017, Telstra’s consumer system couldn’t even support its integration. This led to Telstra having to migrate personal consumer’s to their business platform with wait times of up to 6-months!
This is an area where future-AI could significantly help, by either acting as hundreds of software engineers to assist in migrating to a new platform, or as a replacement for traditional databases altogether.
As Prof. Simon Angus pointed out, the first one to develop this kind of technology would make a lot of money!
The main point which kept coming up at this event is AI works best as an assistant to human-work and not as a full replacement for it. Which sounds logical, however it’s not the narrative AI-companies continuously sprout. Because for them, these models are more highly valuable when they promise to replace people altogether.
Finally and one of the most fascinating points I heard throughout the day was around how AI-tooling would become a vital part of companies going forward. In fact, Prof. Simon Angus suggested that future employees would have this as a requirement when choosing their jobs, forcing companies to compete in this area.
More practically right now, Angus Nalwan from Carsales highlighted the need for companies to put AI through production pipelines to achieve the best outcomes. Ideally, having a holistic approach within the company (every department is on board with this idea) and then measuring it against the metrics of what you’re trying to solve.
Basically, the process should prove AI works for the intended business outcome.
Once again, sounds obvious, but not everyone is doing this!
Of course, in comparison to the rest of this essay, this event focused on AI used within the context of commerce, not Generative-AI to make music, photos or movies.
Potentially, that’s why it was enjoyable too. Because while arguably boring to most, Artificial Intelligence is right at home doing the tasks that bottleneck companies today. Plus, while some jobs will certainly be made redundant, this use of AI will require the use of new roles such as Angus Nalwan’s ‘General Manager of AI’ title at Carsales.
Professor Pitòsh was also keen to point out this would likely negate any cost-benefits of AI-integration, at least in the short term.
In comparison, many AI-startups choose to pit themselves as a replacement for human creativity and artistic output which just feels a little ridiculous. However, it definitely makes for flashier headlines!
WRAP UP
With the AI-race well underway, I want everyone to remember that people are still highly valuable.
Human art and creation is highly valuable.
Despite all the noise around AI, I’ve yet to hear of anyone in my close circle transition from their creative roles into managing purely AI-based pipelines for their work.
Most creatives I speak to loath the idea of AI taking their roles or somehow cheapening their work.
At the same time, despite the many complex issues yet to be resolved (copyright, ethics, environment etc.), I believe AI should not be feared overall. As with any new technology, there are opportunities here to help businesses and individuals thrive.
As Prof. Pitòsh from Monash said:
“Focus on your strengths, not weaknesses. There is something you can do with AI that’s unique to you- find out what it is.”
It’s important to remember the introduction of Photoshop didn’t make everyone a digital artist. Smartphone cameras have yet to make everyone a photographer, and AI will not make everyone a creative capable of producing emotionally moving works.
However, in the context of art, it does increase accessibility. It lowers barriers to entry and makes technically impressive and even personalised outcomes easier to achieve. But at the price of reducing human touch, and therefore perceived value.
I keep focusing on art and creativity because that’s my main area, but I’d argue a painting is generally worth more than a digital print (NFT’s anyone?), a vinyl record is worth more than the digital bits of an MP3, and human creativity will continue to be worth more than AI due to its relative scarcity and desirability.
As mentioned in the previous section, within the context of commerce, AI arguably has more functional applications. Helping companies to achieve technical outcomes previously not possible due to a lack of resources such as time and money.
But make no mistake, AI still requires humans as of right now.
What these technologies can do is incredible, no doubt, and it’s only possible because of the massive advancements in the underlying hardware and human ingenuity that brought it all to life. But these systems should not replace people entirely.
Instead, AI used as an assistive technology can help humans achieve more. Just like the calculator and computer did. It was never about making humans redundant but about increasing efficiency, and therefore truncating our progress as a species.
In an idealistic sense, this means creating better outcomes for everyone across our collective health, economic prosperity and optimism about the future.
At the same time, new technology always comes with a significant drawback. With each new layer of convenience, we no longer have to think as deeply or meaningfully. Remembering surroundings becomes redundant in the age of Google Maps and real-time translation removes the need for anyone to learn a new language when travelling.
Generative-AI will take this to a new level, attempting to make critical thinking a skill of the past. This has many further wide-ranging implications that are too vast to dive into here.
However, in the meantime, I urge everyone to think deeply as these new technologies hit the market.
What am I being sold here? Can this product actually do what it promises? What am I missing out on by not doing this myself or instead bringing on other talented people?
Is this robbing me of my own human experience and therefore growth?
I believe it’s time we break through the smokescreen these big tech companies are trying to cover us in.
AI isn’t magic, it’s technology.
It’s time we treat it as such.
Disclaimer: No AI was used in the making of this piece. It’s still possible!
Related Posts
