Time for some big ideas.
Fast Company with “A new IT maintenance business model”. A new business model for IT maintenance could significantly reduce carbon emissions and costs associated with outdated server replacement practices. Currently, companies are replacing servers not due to failures but because maintaining them becomes prohibitively expensive after three to five years, leading to unnecessary waste and emissions. According to Dell’s lifecycle assessment, manufacturing processes account for up to 83% of a server’s carbon footprint, making it crucial to extend hardware lifespan. If just 10% of the global server market, which ships around 12 million units annually, were to extend their lifespan from four to six years, it could potentially save 5.8 million tons of CO2 emissions, equivalent to the annual impact of 630,000 households. This shift not only benefits the environment but also presents a cost-saving opportunity, estimated at $650,000 to $850,000 for a mid-sized company with 200 servers.
I’m a touch fascinated by Enshittification, and Wired writes how Cory Doctorow’s theory of “enshittification” suggests that technology platforms can deteriorate from within as they prioritize profit over user experience. This concept raises concerns as artificial intelligence systems become more profitable and powerful, potentially leading to a similar decline. In a recent example, a user shared how a recommendation from an AI model led to an excellent dining experience, highlighting the complexity and impressive nature of the AI’s decision-making process. However, as AI continues to develop, experts warn that the industry must avoid the pitfalls of enshittification to ensure that artificial intelligence remains beneficial for users rather than succumbing to profit-driven deterioration. Ars Technica follows up, highlighting how Doctorow argues that this decline is a structural issue affecting major platforms like Facebook and Google, driven by the erosion of competition, government regulation, interoperability, and labor power. He suggests that restoring these constraints is essential for reversing the trend and improving the quality of online interactions.
I’ve highlighted OpenAI’s financial challenges. Ed Zitron does that to Anthropic. Anthropic has reportedly spent over $2.66 billion on Amazon Web Services through September 2025, exceeding its estimated revenue of $2.55 billion during the same period. This financial strain highlights the unsustainable costs associated with operating generative AI services. According to sources, Anthropic’s substantial expenditures include $1.35 billion for AWS in 2024 alone, illustrating a trend where operational costs rise in tandem with revenue. The company’s largest customer, Cursor, also faced a dramatic increase in AWS costs, rising from $6.19 million in May 2025 to $12.67 million in June 2025, as a result of new pricing strategies introduced by Anthropic.
Why do we care?
Let’s throw some big questions on the table.
Why are we still tossing out servers every three to five years when keeping them longer saves money and cuts emissions? Fast Company’s right — the waste is built into the business model. Maybe MSPs should profit from maintaining, not replacing.
Then there’s Doctorow’s “enshittification” theory. Platforms rot when profits come before users — and that’s the road AI’s already on. Are we building our businesses on platforms that’ll turn hostile once they’ve got us hooked?
And Ed Zitron’s digging into Anthropic’s books should make everyone nervous — they’ve spent more on AWS than they’ve made in revenue. If the biggest AI vendors can’t make money at scale, what does that say for everyone downstream?
So here’s the challenge: Are you chasing the shiny thing, or building something that lasts? Sustainability — environmental, economic, and technological — isn’t the opposite of innovation. It’s what keeps innovation alive.