3+ hour, 6+ min ago  (1765+ words) Foundation models are now mainstream. We first saw them in language, then vision, and now also in video and speech. The recipe by now is familiar: first, pretrain a big neural net on large enough data, then apply the model…...

22+ hour, 36+ min ago  (1449+ words) A retrospective on my MS thesis, the leaderboard it placed on, and the LLM shift that has reshaped the field since. In March 2024, I submitted my MS thesis on Emotion Recognition in Conversation (ERC). The model, Emo Net, achieved a…...

3+ day, 6+ min ago  (1243+ words) Why your AI model can be wrong with 99% confidence Last year, I was feeling a bit whimsical on a Saturday and decided to ask Chat GPT a fairly simple question: "Who won the Nobel Prize in Physics in 2025?" Chat GPT…...

3+ day, 21+ hour ago  (1040+ words) A beginner's honest walkthrough of Extract, Transform, Load using the Git Hub API But it also came with pressure. Suddenly this wasn't just a personal goal I could quietly abandon if things got hard. People were watching. People were in…...

1+ week, 1+ hour ago  (1412+ words) Exploring one of the most overlooked bottlenecks in QML: getting data into a quantum computer efficiently. This may sound simple at first, but in practice it is surprisingly difficult. As the size and complexity of the data increase, the cost…...

1+ week, 22+ hour ago  (1227+ words) A practitioner's warning about generated variables in causal analysis An analyst joins LLM-extracted themes from a call corpus to the customer table. Customers without transcripts get NULL. NULL gets filled with zero, or with "no issue mentioned," or quietly omitted…...

1+ week, 1+ day ago  (985+ words) If you don't want to be left behind, start doing these things with Claude Kids nowadays don't know what it's like. I used to spend hours: Even in just the last year, as AI tools have become increasingly more advanced,…...

1+ week, 1+ day ago  (1841+ words) Whenever you can rewrite a (stochastic) optimization problem so that fixing some variables makes the rest separable, you could try Benders. In my first TDS post, I wrote about translating a real-world problem into an integer linear program. In my…...

1+ week, 3+ day ago  (958+ words) Enterprise knowledge graphs have become the most widely used business semantic layer, providing a unified view of an organization's suppliers, contracts, products, partners etc. As a result, they evolve organically over time to become very large, with millions of nodes…...

1+ week, 4+ day ago  (1170+ words) 95% of enterprise AI pilots fail to launch. Why? If you have spent any time in enterprise AI over the last two years, you know the pattern. A small team builds a proof-of-concept using a state-of-the-art Large Language Model (LLM). The demo…...