The Economics of Artificial Intelligence, Causal Tools, ChatGPT’s Impact, and Other Holiday Reads
Feeling inspired to write your first TDS post before the end of 2024? We’re always open to contributions from new authors.
Our guiding principle is that it’s never a bad time to learn new things, but we also know that different moments call for different types of learning. Here at TDS, we’ve traditionally published lots of hands-on, roll-up-your-sleeves guides and tutorials as soon as we kick off a new year—and we’re sure that will be the case come January 2025, too.
For now, as we enter the peak of the holiday season, we wanted to highlight some of our best recent articles that call for a bit more reflection and a slower pace of processing: stories you can savor as you lounge on a comfy armchair, say, rather than while typing code away on your laptop (though you can do that too, of course; we won’t hold it against you!).
From the cultural impact of AI-generated content to a Bayesian analysis of dogs’ pooping habits (yes, you’ve read that right), we hope you enjoy this lineup of thought-provoking, engaging articles. And stay tuned: we can’t wait to share our 2024 highlights with you in next week’s final-edition-of-the-year Variable.
- The Economics of Artificial Intelligence — What Does Automation Mean for Workers?
In his comprehensive analysis of AI’s effect on the workforce, Isaac Tham introduces a powerful framework: “AI augments or automates labor based on its performance relative to workers in a given task. If AI is better than labour, labour is automated, but if labour is better than AI, AI augments labour.” He goes on to unpack the stakes, risks, and potential benefits of AI’s rapidly growing footprint. - The Cultural Impact of AI Generated Content: Part 1
Business implications take up much of the space in conversations around AI, but as Stephanie Kirmer stresses, we shouldn’t ignore the potentially seismic shifts AI-generated content causes in the cultural sphere, too: “It would be silly to expect our ways of thinking to not change as a result of these experiences, and I worry very much that the change we’re undergoing is not for the better.” - ChatGPT: Two Years Later
November 2022, when OpenAI launched the chatbot that would change everything (or at least… a lot of things), feels at once like two days and two decades ago. To help us make sense of our post-ChatGPT world, Julián Peller presents a panoramic overview of the past two years, a period of monumental transition within the “generative-AI revolution.” - The Name That Broke ChatGPT: Who is David Mayer?
For anyone who enjoys their explorations of AI’s inner workings with a generous dose of intrigue and mystery, Cassie Kozyrkov’s latest article fits the bill: it tackles some of the thorniest questions around LLM-based tools (privacy, bias, and prompt hacking, to name a few) through the example of one elusive name. - Overcoming Security Challenges in Protecting Shared Generative AI Environments
Approaching the problem of security in AI products from a different angle, Han HELOIR, Ph.D. zooms in on the particular challenges of multi-tenancy—the increasingly common situation when different groups of users (like multiple teams within a company) rely on the same data and LLM resources.
- Understanding DDPG: The Algorithm That Solves Continuous Action Control Challenges
Why not take the time this holiday season to expand your knowledge of deep reinforcement learning algorithms? Sirine Bhouri’s debut TDS article walks us through the theory and architecture behind the Deep Deterministic Policy Gradient (DDPG) algorithm, tests its performance, and examines its potential applications in bioengineering. - LLM Routing — Intuitively and Exhaustively Explained
With thousands of large language models to choose from, how should practitioners decide which ones to choose for a given task? Daniel Warfield’s accessible deep dive into LLM routing explains how this “advanced inferencing technique” streamlines this process and how the different components it relies on complement each other. - The Intuition behind Concordance Index — Survival Analysis
Understanding and preventing churn remains one of the most common goals for industry-embedded data scientists. Antonieta Mastrogiuseppe provides a thorough primer on the underlying math of survival analysis, and the key role the concordance index plays in assessing a model’s accuracy. - Dog Poop Compass
Can a 5-year-old Cavalier King Charles Spaniel teach us important lessons in Bayesian statistics? It turns out the answer is yes — as long as you follow along Dima Sergeev’s gripping account of his attempts to detect patterns in his dog’s “bathroom” rituals. - Causality — Mental Hygiene for Data Science
To round out our lineup this week, we invite you to dig into Eyal Kazin’s thoughtful reflection on causal tools—and when (and whether) to use them. Based on his recent PyData Global conference lecture, this article balances a big-picture analysis of causal inference with the nitty-gritty factors that shape the ways we apply causal thinking in day-to-day workflows.
Thank you for supporting the work of our authors! As we mentioned above, we love publishing articles from new authors, so if you’ve recently written an interesting project walkthrough, tutorial, or theoretical reflection on any of our core topics, don’t hesitate to share it with us.
Until the next Variable,
TDS Team
The Economics of Artificial Intelligence, Causal Tools, ChatGPT’s Impact, and Other Holiday Reads was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.
from Datascience in Towards Data Science on Medium https://ift.tt/N2iF6rC
via IFTTT
