Related Posts
Community Builder
Bowl Leader
Conversation Starter
Bowl Leader
Welcome to your new bowl, fellow 🐠! Go wild.
Additional Posts in Data & Analytics Consultants
Bowl Leader
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
GPT-3 can also write memos, emails, summaries, and maybe even slides once someone creates a good encoding. So really all of our jobs are equally at risk from GPT-3, which is that it'll probably automate away the menial parts of our jobs leaving us to focus more on the thinking and problem solving parts.
Prepare for even more boilerplate RFPs
Subject Expert
Far from it. Majority of our time goes into Problem understanding, formulation into mathematically solvable approaches, getting data ready for ingestion into the approaches and then at last training and deploying the model.
It not like we will say “Go fetch data, compute features, train model and improve it” and GPT is like, “Sit right and enjoy your coffee 🤣”
So, it is pretty cool but far far away from replacing us.
Thank you for educated answer. Most colleagues are not realizing, How ML and AI works, especially training aspect of model.
Remember when RPA was supposed to replace traditional workers?
Literally every library/framework is meant to automate coding if you think about it. Yet the need for programming is still at an all time high
Coding is never going to be “less of a competitive advantage” coz whatever code GPT-3 is writing, someone needs to understand it to maintain, debug or make any changes to it in the future.
But, I feel everybody needs to go “cloud” ASAP. Almost every company is gonna ask for atleast one cloud certification in the near future.
In order to replace programmers with robots, clients will have to explain their requirements clearly. I think our job is safe
Lmao says a “strategy” consultant ..
This banter was not of quality
Whatever AI or module that is writing the code will need maintenance.
Can anyone think of reasons why corporations with proprietary information/data/technology may not want some AI system created elsewhere to have free reign and access to everything on its servers?
Nope sounds like a winner to me
Programming has continually gotten easier, which lets us be more productive. The switch from machine language to high level code made coding much easier but it doesn’t seem to have made developers obsolete.
I work for big tech and this already happened, data scientist are no longer cleaning data or doing data engineering, building pipelines, or training and validating models.
All the models are already built, the data scientists role is to apply the correct model and build presentations to communicate impact and value.
The product data scientists are basically just writing sql all day and trying to find statistically significant trend. Overall the job is easy but the standards are high.
Bowl Leader
I've built something like this for a repetitive project, where many similar models from a parametrized pipeline were needed for production use. The framework was also generalized whenever possible, making it feasible to scale modeling efforts. However, as a data science team, we had plenty of other work. The separate development of both new models and framework tools never stopped, nor could it. Too many different use cases.
Making oneself indispensable means locking your client in. Nuclear family notwithstanding, I don’t think most people want that in their personal relationships. Why would it be different in professional ones?