By Joe Triccas Product Manager at This is Milk Self-sovereign data ownership may sound like a complicated concept, but it’s quite straightforward. Let’s consider a scenario: you’re on a weekend trip, and you suddenly fall ill. You make an appointment with a doctor, but they can’t access your medical records because they are owned by the NHS and not directly accessible across trusts. This means they are flying blind, hoping that you remember all your allergies and medical history accurately. (The NHS has taken great steps to improve this, alas the issue continues internationally) In the current data ownership model, you can request your medical records from the NHS, but they are redacted and not entirely under your control. Doctors often trust the data more than what you say, which can create issues in medical treatment, especially for us in the above scenario. However, imagine a world where you own your medical records. They are stored securely on your phone, protected by biometrics like your face scan or fingerprint. If you travel and fall ill, you can grant temporary access to your medical records to the doctor, who can access what they need. You get to see what information they access, and everyone is happy. What is more, in a world of self-sovereign data ownership, you can offer your medical data for research and be compensated directly for it. This creates a more equitable system where individuals benefit from their data, rather than companies reaping all the rewards.
This model of data ownership is not limited to medical data. It can be applied to any data that individuals generate, including their online activities, financial transactions, and personal preferences. By allowing individuals to own and control their data, self-sovereign data ownership can create a fairer, more transparent, and more equitable system of data ownership. Furthermore, self-sovereign data ownership can facilitate the coexistence of humans and advanced AI systems. As AI systems become more advanced, they will require vast amounts of data to operate effectively. However, the current data ownership model creates significant barriers to the development of advanced AI systems. Companies are often unwilling to share data with each other, and individuals are reluctant to share their data because of concerns about privacy and security. In fact, most models thus far have essentially been built under the ‘public data’ approach i.e. it will just scrape data from the entire internet, as long as it doesn’t have to hack anything to get the data, then it’s fair game. It is also worth noting that currently, the companies that train AI models own that AI model. I think it is fair to say the case of ‘owning self-aware conscious beings’ has been asked and answered and it would be a real shame, if not morally unacceptable, were we to repeat the mistakes of the past when birthing this new form of consciousness. Self-sovereign data ownership is a simple but powerful concept that has the potential to transform how we interact with our data. It can create a fairer and more transparent system of data ownership, where individuals have control over their data and can benefit from it. It can also facilitate the development of advanced AI systems and create a more equitable relationship between humans and AGIs. It is time for individuals, companies, and governments to recognise the importance of self-sovereign data ownership and work towards implementing it. In my mind, I see the current attitude towards personal data as being very similar to the early industrial revolution approach to ecological protection – if you can do it just do it and worry about the consequences later. Centralised data stores is indeed more ‘efficient’ from a pure data storage perspective, however, comes with many drawbacks, the biggest of them being, if the central data store is hacked, everything is generally hacked. The case for self-sovereign data is strong, however, is unlikely to come to fruition as companies simply do not profit from it. It is going to take a motivated, informed population to demand that this is the direction we head in, protecting us from a potentially deadly outcome as AI continues on its current trajectory.
0 Comments
Leave a Reply. |
Archives.
October 2024
Categories.
All
|