The first four years of the generative AI revolution have tended to see power flow in one direction. Big Tech firms unleashed their tools on the planet and expected users to like it or lump it. Unhappy with the amount of data collected about you in the pursuit of convenience from your chatbot? Tough luck. Squeamish about the provenance of all the data AI systems have been trained on, and whether it was legally obtained? Not a lot you can do about it.
But as the technology continues to advance, and users become more comfortable with its capabilities, that imbalance is starting to shift. And Big Tech might regret what it’s wrought. If the first few years of the generative AI revolution were about vendor lock-in and AI companies potentially exploiting intellectual property they had no right to use, the next few could see David start to fight back against Goliath.
Like many others, I bought a digital photo frame from Amazon this week, inspired by developer Wes Bos’s reverse engineering of a $50 photo frame he bought from the retailer and turned into a smart display. Within a couple of hours, and with the help of Claude, I’d sidestepped the built-in software designed to lock me into a recurring subscription just to view my own photos, installed an operating system of my choice, and repurposed the frame into a wall-mounted display of regularly updating headlines to keep me on top of the news agenda. I’ve also resuscitated a decade-old printer for which there are no compatible drivers for modern operating systems, and overridden proprietary software for a thermal receipt printer to turn it into an on-demand, tear-off to-do list.
Next up on that to-do list? Sic a swarm of AI agents on the locks that prevent 20 e-ink price tags, once mounted on supermarket shelves, from displaying what I want, and see if they can crack them. Then hack together a hardware system to power the tags, replacing the rail-based system that once sent them signals.
I’m not alone. Want a photo editor or graphics package like Adobe Photoshop but don’t want to pay the exorbitant monthly subscription for it? Folks have used AI to reverse engineer the way the app works and vibe-code alternatives of their own. Same with Word and any number of other apps. Want to play console games in a web browser? Sure. AI has helped with that, too. One of the hottest GitHub repositories right now is Reverse Engineer Anything (REA), which aims to make the process easier.
The maker movement has always existed, but AI’s ease of use helps smooth over the bumpier elements of getting into that space. (You can, for instance, now send a photo of a circuit board to ChatGPT and ask it to identify what does what, rather than pore through a manual yourself or undertake hours of trial and error with a multimeter.) But what’s even more interesting than people tinkering and tweaking is the glee with which many of us are doing it.
There’s a sense of moral rectitude that underpins much of the current trend of reverse engineering anything. Yes, AI makes it technically possible to do a lot of this stuff far more easily than before, but it’s the burning sense of longstanding injustice, of being screwed over by endless subscriptions and the inability to ever own anything, of being bilked and badgered for more money, that is fueling the movement. I take glee in giving the finger to an exploitative photo frame app and avoiding the risk of installing malware on my phone just to use something I bought. I feel morally justified in doing so because too many tech companies, big and small, have screwed me over before, so screw them. And better still, I have control over what happens to my device next.
Depending on the jurisdiction, some of this reverse engineering may not be legal, though consumers also have rights to tinker with what they buy. But many of us reckon we’re just paying back a fraction of what’s been done to us before. And if tech companies didn’t have any qualms about what they did for decades, why should we?