Mine or the Collective Experiences?

Tyna Hope
3 min readMay 31, 2024

--

In a world where all work seems to be able to be done by machines, what will we be able to offer?

It seems, ATM, that my most valuable asset is my internet trail. My past or, you could say, my experience. Those snippets of data that I leave in my wake to be used as building material for the hungriest creatures we have seen to date; AI models.

From a philosophical perspective, this is interesting to me because there is only so far that we can expect to go back, since the internet wasn’t born until the 1950s and not commonly used by citizens until 1990s. While it is possible to upload artifacts to websites, these are lacking the behavioral information that lead to those artifacts, and the behavior is often what we want to predict.

At first glance it seems that anyone’s continued contribution to this body of knowledge used to train and validate all kinds of models should be encouraged, perhaps even rewarded. However as of January 2024, there are over 5 billion users of the internet. So, each of us are not really that important to a system that mines all of that information and is designed to generalize on all of our data. Generalizing is something that reflects the most represented group in the data set. In the basics of statistical concepts, as long as your sample has the same distribution as your target population, then just using a fraction is often considered good enough. Perhaps this is why companies that need our data for their business model provide such a small reward for you to share your data. It is often trivial or non-existent. Or worse, we are training their models under the guise of security protocols, such as re-captcha, and so we are more than data providers but also model validators.

One could argue that the most interesting data are those being gathered right now. In areas such as consumer trends, investing, and outbreaks of diseases, to name just a few, the most interesting is the now compared to averaged trends over longish periods of time. The question here becomes what constitutes “long enough” to get an informative average that tells us something interesting has just happened. Our collective memories are short, and our AI tools are being developed with memories that are equally so.

So if my behavior and creative ideas are useful for the AI future, why does everything now seem to expect that we relinquish our creative thoughts in favor of using the collective wisdom served up by LLMs and GenAI? Even in relationship creation. For example during a job search we are expected to deliver the perfect version of the cover letter, outreach email, and LinkedIn connection request. Likely all of these were generated with the help of AI, or at least we are presented with the tools and strongly encouraged to use them. We have been fed the idea that what we have to offer as authentic individuals is insufficient and that these tools will deliver better, faster, and without us needing to emotionally invest in the results.

Photo by Dari lli on Unsplash

It seems that the world is more and more connected and that the latest batch of technology is intended to help us deliver even more and faster. But, I wonder if we haven’t lost something. I do use social media, I have no choice but I would rather reach out to fewer people with conviction, and make more genuine connections while using my own imperfect voice.

The feedback loop of AI use on humanity behavior has yet to be fully explored. There are problems that we have not yet solved, and that desperately need creative energy. If we are always looking at a tool that gives us the average or best fit of past experiences, we may not get the desired result, rather we will get what was done before. We need to find a way to keep human generated ingenuity in the loop or we will be doing the same past actions and expecting a better result. As someone once said, that is the definition of insanity.

--

--

Tyna Hope
Tyna Hope

Written by Tyna Hope

Electrical Engineer who worked as a data scientist then as a product manager, on LinkedIn. Opinions expressed are my own. See Defy Magazine for more: defymag.ca

No responses yet