Who should I follow to learn about training data?

If you mean AI/ML training data, I’d follow a mix of researchers, practitioners, and people focused specifically on data quality, datasets, synthetic data, and evaluation.

To learn about Training Data - ranging from data curation and data quality to synthetic data generation and dataset scaling for machine learning and LLMs - you want to follow researchers, data engineers, and organizations that specialize specifically in how data shapes AI.

Here are the key people, researchers, and thought leaders to follow, grouped by their specific focus areas:

1. Data Quality, Reliability, and Curation Experts

• Barr Moses: Co-founder and CEO of Monte Carlo. She is a leading voice in data observability , data quality, and reliability. Great to follow for understanding how to catch errors and bad pipelines before data hits models. Top 22 Data Influencers to Follow in 2025 - Rivery

• Joe Reis: Co-author of Fundamentals of Data Engineering and host of the Monday Morning Data Chat. He focuses heavily on the engineering side of preparing and processing data for analytics and AI. Top 22 Data Influencers to Follow in 2025 - Rivery

• Ben Rogojan (The Seattle Data Guy): Known for highly practical breakdowns of data pipelines, architectures, and the infrastructure required to feed machine learning models effectively. Top 22 Data Influencers to Follow in 2025 - Rivery

2. AI Dataset Pioneers & Visionaries

• Dr. Fei-Fei Li: Co-Director of Stanford’s Human-Centered AI Institute. She is legendary for creating ImageNet , the foundational dataset that essentially kickstarted the modern computer vision and deep learning revolution. 12 Top AI Experts and Influencers Every IT Leader Should Follow

• Timnit Gebru: Founder of the Distributed AI Research Institute (DAIR). She is a prominent voice on data ethics, dataset documentation (such as "Datasheets for Datasets"), bias, and the societal impacts of training data.

• Margaret Mitchell: Chief Ethics Scientist at Hugging Face and former co-lead of Google’s Ethical AI team. She focuses deeply on dataset transparency, evaluation, and safety data.

3. LLM Data Scaling, Synthetic Data, and Fine-Tuning

• Andrej Karpathy: Former Director of AI at Tesla and co-founder of OpenAI (now building Eureka Labs). His technical breakdowns and tutorials on neural networks, tokenization, and LLM data mechanics are widely considered masterclasses for practitioners. Top 24 Generative AI Influencers to Follow on Social Media in 2024

• Sebastian Raschka: Staff Research Engineer at Lightning AI and author. He regularly publishes deep, accessible breakdowns on how to build, fine-tune, and curate data for Large Language Models. Top 20 AI Educators You Should Follow in 2026 - SocialBook

• Hugging Face (The Organization & Core Researchers): While not a single person, following key researchers and community builders at Hugging Face (like Thomas Wolf or data-focused teams) gives you a front-row seat to open-source dataset curation, benchmarking, and data processing libraries.

Key Substack Newsletters & Communities to Watch

• The Batch (DeepLearning.AI / Andrew Ng): Frequently covers shifts in data-centric AI (moving away from just scaling models to engineering better data).

• Data-Centric AI Community: Look into initiatives and discussions spearheaded by Andrew Ng’s circle, which emphasize that cleaning and improving training data is often more impactful than tweaking model architectures.

Are you looking to learn more about a specific angle of training data - such as data engineering pipelines, LLM fine-tuning datasets, or data cleaning tools?

Learning about training data - the core fuel for machine learning and artificial intelligence - is best done by following leading researchers, data-centric AI pioneers, and organizations specializing in data curation, annotation, and engineering.

Top AI & Data-Centric Experts to Follow

** Andrew Ng** : Founder of DeepLearning.AI and co-founder of Coursera, Ng has heavily championed a "data-centric AI" movement. Rather than obsessing purely over model architecture, he frequently shares practical insights, tutorials, and frameworks on how improving training data quality fixes broken AI behavior.

Andrej Karpathy : Former Director of AI at Tesla and co-founder of OpenAI, Karpathy is renowned for his brilliant, accessible educational content on neural networks and large language models, breaking down how raw data turns into functional intelligence.

Fei-Fei Li : Co-director of Stanford’s Human-Centered AI Institute and the creator of ImageNet - the landmark dataset that helped spark the modern deep learning revolution. Her work and writing offer foundational perspectives on the importance of large-scale visual and multimodal training datasets.

Ian Goodfellow : The inventor of Generative Adversarial Networks (GANs), Goodfellow's research is essential for understanding how models can learn from distributions, generate synthetic training data, and handle unsupervised challenges. Top 10: Machine Learning Leaders | AI Magazine

Industry Leaders and Data Platforms

• • Scale AI : A premier data platform for AI that regularly publishes research, reports, and engineering blogs regarding human-in-the-loop data curation, RLHF (Reinforcement Learning from Human Feedback), and fine-tuning datasets for frontier models. How I Would Learn to be a Data Analyst

• Hugging Face Datasets : Following the official Hugging Face channels and community updates gives you a direct, hands-on window into open-source training sets, how they are structured, tokenized, and evaluated.

• Kaggle : The community and educational arm of Kaggle offers public datasets, notebooks, and discussions where data practitioners share real-world data cleaning, exploratory analysis, and feature engineering techniques. Best Resources to Learn Data Analytics in 2025

If you want to narrow this down, let me know:

• • Are you interested in computer vision data , text/LLM data , or multimodal data?

• Are you looking for academic theory or hands-on data engineering tutorials?

If you mean the “Training Data” AI podcast, follow host Sonya Huang of Sequoia Capital; its guests include AI researchers and founders such as David Silver and Lasse Espeholt.(music.amazon.com)(podcasts.apple.com)

If you want to learn about training data, follow leading academics, practitioners, and thought leaders who specialize in data science, AI ethics, and analytics strategy. Some of the most influential voices are connected to MIT, Wharton, and Oxford programs, as well as independent experts shaping how organizations use and govern data.

📌 Key People & Institutions to Follow

🔑 How to Engage With These Experts

• LinkedIn & Twitter/X: Most of these leaders share frequent updates, research, and case studies.

• Executive Programs: Enrolling in MIT, Wharton, or Oxford executive education programs gives direct access to instructors and networks.

• Webinars & Conferences: GTLI symposiums and eLearning Industry webinars are excellent for staying current on training data applications.

• Publications : Look for white papers, blogs, and research articles authored by these experts - many are cited in AI training datasets themselves.

⚠️ Risks & Considerations

• Hype vs. Substance: Some “training data” discussions online are buzzword-heavy. Following credible institutions ensures you get practical insights rather than marketing fluff.

• Ethics & Bias: Training data is central to AI bias issues. Experts like GTLI emphasize credibility and trust, which are critical to avoid misinformation.

• Rapid Change: AI models and training data practices evolve quickly. Staying connected to academic programs and thought leaders helps you keep pace.

👉 Would you like me to curate a list of specific social media accounts (LinkedIn/Twitter) for these experts so you can follow them directly? That way, you’ll have a ready-made feed of training data insights.