AI and the Misconception of "One Big Thing"

AI and the Misconception of "One Big Thing"

by Timothy Coleman - 10/Apr/2025

The Misconception About AI (and Blockchain)

I recently read an article titled Bored of It by Paul Robert Lloyd, which, despite not explicitly mentioning AI, sparked some thoughts about how we discuss technology - particularly artificial intelligence (AI) and blockchain.

Something I've noticed about the discourse surrounding these technologies is how frequently people talk about them as singular, monolithic entities. People say, "the blockchain" or "AI did this," suggesting there's one large, unified system behind everything.

This reminds me strongly of when blockchain first entered mainstream consciousness. Everyone began referring to "the blockchain" as if it were one giant network or singular database.

My Experience with Blockchain

I tend to engage with technologies during the innovation or early adopter stages. While I wasn't involved with AI at its innovation stage, I was actively learning during the early adoption phase. However, with blockchain, I entered at the innovation stage around 2015 or 2016 - back when Bitcoin involved downloading a 3GB blockchain via a QT app on a desktop.

The insights were valuable. As blockchain moved through early adopter stages to mainstream usage, the misconception of it being "one thing" came into being.

AI: Not One Giant Computer

This misconception is happening again with AI. Many speak about AI as if it’s one vast supercomputer performing every task - creating videos, images, and texts, and learning all about us. While large, centralised systems (like ChatGPT by OpenAI) exist, AI is fundamentally software. It can be downloaded, run offline, and decentralised.

For instance, I frequently use AI tools for voice-to-text transcription directly on my laptop. These offline tools outperform alternatives like Apple's built-in dictation, demonstrating AI's versatility and local applicability.

Privacy and Efficiency with Local AI

Using offline AI tools also addresses two significant concerns: privacy and resource consumption. I'm cautious about sending voice samples and personal data to centralised AI services, preferring local processing for privacy reasons.

From an energy and resource perspective, local processing is vastly more efficient. Running voice transcription software offline only requires my laptop’s existing resources, avoiding the massive energy consumption and cooling demands of centralised data centres.

Blockchain as Decentralised Software

Similarly, blockchain technology isn't a singular entity but software that anyone can deploy independently. I can easily create a private blockchain and grant access to just ten people if I choose. Blockchain, like AI, is replicable, distributable, and localisable.

Potential Societal Benefits

Despite common misconceptions, both AI and blockchain hold immense potential to benefit society, particularly in reducing repetitive administrative tasks. General Practitioners, for instance, spend significant time on administrative work rather than their specialised medical duties. AI can alleviate this workload substantially.

As a freelancer, I personally rely on AI tools to handle administrative tasks, which helps me spend more time on client work - tasks that directly generate revenue. While I'd ideally prefer to hire assistance for administrative tasks, financial limitations make AI tools invaluable.

Future Directions

Looking ahead, I anticipate more localised processing with AI, helping mitigate energy consumption and enhancing privacy. Organisations and individuals alike should embrace this decentralised model, using technology responsibly and efficiently.

Conclusion

Both AI and blockchain technologies offer incredible utility and versatility, but we must discard the myth of them being singular, monolithic entities. Understanding and embracing their decentralised nature is essential for leveraging their true potential effectively and responsibly.