David Culley@davidculley@hachyderm.io
#Introduction
I’m a data engineer (previously a ML engineer) and I’m highly critical of the unethical habitat-destroying use of AI.
More specifically, I'm an analytics engineer, something in-between a data engineer and a data analyst. I make sense of the raw data that my co-workers collect in the sense that I shape the raw data into business concepts like "customers" or "orders" or "payments" that are actually useful. The pure data engineers talk too little with stakeholders and have too little domain knowledge to do that and the analysts who I enable with that work lack the technical capabilities to do that (in the same quality I can do it). I write Python and SQL code. My job is to create a maintainable data architecture, ensure the reliability and trustworthiness of the data, and save data analysts from presenting flawed analyses and insights to executives who'll make decisions based on those analyses based on my shaping the data.
I come from sigmoid.social that I joined in Nov 2022, since most people due to whom I joined that instance quickly returned to X thereafter. Apart from GenAI art spam, that place seems largely inactive. It’s not the place I’d hoped it would be, and I don't feel at home there with my critical views of planet-boiling and water-sucking LLMs and generative AI.
I originally studied to become a ML engineer to, for example, use ML to translate the communication of sperm whales and do something good, but I realized that in today's context—GenAI devaluing work—I'd only be a useful idiot for those wealthy folks who want to access skill while denying skill the access to wealth.
I still believe in the uses of "classic" machine learning (from the pre-LLM era), at least for now, but even the interest in working in that area is dwindling with all these ignorant ML acolytes and techbro fanboys of billionaire CEOs that drove me away to data engineering, and with how these pattern recognition technologies are more and more being abused by authoritarian regimes worldwide (face recognition to persecute minorities and protestors, to give an example).
I realize that organizing data could also be considered problematic in context of ad-tech capitalism, but that skill can be used for sth other than surveillance.