I promise I'm Human

I promise I’m a human. But should you trust me? In the past readers might scoff at the idea that a computer could effectively imitate a human to the point where flesh and bone writers have to plead with their audience to believe them. But with the recent upsurge in content created by ever more convincing generative AI tools, what on earth can anyone believe any more?

It’s an important question, to state the obvious, and it ranges from the banal, to the concerning, to the downright terrifying; and that’s before we even start to consider AGI and its implications for humanity. On one end of the scale we have meme worthy pictures of the pope wearing a puffer jacket (it turns out he prefers the more traditional look) and on the other, are formidable large language models capable of imitating humans to such a degree that scam emails, text messages, and even voice calls become indistinguishable from genuine human communication.

When you add in the influence of social media and our already divided society, the implications of widespread digital fraud, capable of destroying lives and breaking down social cohesion, are unpleasant to put it mildly.

The creators of the tools causing these concerns are well aware of the issues, and some are already busy finding ways to tackle them. One example is Worldcoin, a project supported by Sam Altman, the CEO of OpenAI, the company behind ChatGPT.

Worldcoin might sound like one of those many digital token ventures that popped up during the cryptocurrency craze, and in a way, it is. But its objectives are closely tied to the rise of generative AI. Essentially, Worldcoin aims to offer two things: a special identifier for "personhood" and a token that verified individuals can earn, use, and have a say in how it's used. To prove their identity, users employ an "orb" to scan their iris, creating a unique identifier stored on the blockchain. This identifier can then be used to verify their human status in any digital scenario, whether it's texting your mum to ask for some cash, voting in the next general election or claiming UBI (in Worldcoin tokens of course)

While the goals of this project are relevant and important in today’s technology climate, there’s a few issues that are important to address. First off, privacy. Small orbs working their way around the world to scan as many eyeballs as possible in order to prove humanity is almost as sci-fi as it gets, and the data that they will collect will be enormous. Sure, there are guardrails in place and the blockchain implementation is meant to be secure and immutable, but the public’s trust in corporations and their use of data is pretty low. Second, in much the same way that screenshotting NFTs became an online joke, the methods of proving someone’s integrity of humanhood could well be easily copied by fraudsters on both a small and large scale. A bank may have the capacity and capability to interrogate a blockchain to prove their applicant is human and unique, but I’m not sure my mum will be doing so to check the whatsapp she received with a code at the end is genuine.?

Moreover, establishing personhood is just a piece of the puzzle when it comes to addressing a world where AI generates digital content. Simply tagging my identity onto something I've created only confirms that it was made by a human, but it doesn't guarantee its accuracy or verify that I haven't used a generative AI tool to produce misleading content that spirals out of my hands, takes on a life of its own, and leads to unforeseen outcomes.

Work will continue on how to both prove personhood and identify content that has been created by AI. However it's arguable that there will be one solution to the problem. Just as generative AI is working on the ability to be multimodal and comprehend and create diverse types of content, the measures to keep it in check and avoid the worst case scenario must also be diverse and comprehensive. Personhood verification will be combined with regulation, AI content identifiers, watermarking, and traditional human verification. By employing this range of approaches, we can ensure effective control and oversight of generative AI while maintaining a balance between innovation and responsible use.

DISCLAIMER: This article was NOT written using ChatGPT (but I used Midjourney to create the image!)