AI Office Hours

After our last event, the feedback was clear: folks are really interested in talking and learning more about LLMs — and how we can bring them into our own workflows. But before we get to using them well, we need a solid understanding of how they actually work.

So this session goes back to the roots. We'll start with how machines first learned to handle sequences — recurrent networks, LSTMs, and GRUs — and the tricks they used to "remember" context, plus where they hit their limits. From there we'll follow the breakthrough that changed everything: attention and the Transformer, and how that architecture opened the door to the GPTs and LLMs we all rely on today.

• The goal is intuition first, plain English, with enough under-the-hood detail to make the "oh, that's why it works" moments click. No deep learning background required — just curiosity.
• Once we've got the foundation down, future sessions can dig into the fun part: actually putting these models to work.

Feel free to drop any questions about AI in the comments below — we'd love to fold them into the discussion!

Each session, we take what actually happened in AI this week and make sense of it together — separating real shifts from press-release noise.
Not a one-way walkthrough. Here's how it runs:

• Story of the week — I bring 2–3 of the biggest developments and we unpack what they really mean, past the hype.
• Hot or not — live poll. I put a headline on screen, you vote: important or overblown? Then we argue.
• Bring your own headline — saw something that confused or excited you? Drop it in chat and we'll dig in as a group.
• Open floor — questions, predictions, disagreements all welcome.

Beginners leave understanding why the week's news matters and how it fits together. Folks already in the field get a sharp, opinionated read and a room to test their takes. Either way, you'll be able to hold your own in any "did you see what just dropped?" conversation.
No prerequisites. Bring a question — or a headline. Online via Zoom.

One ask: share our group on LinkedIn and other social media, and with friends — the more people, the better the conversations. 🙌

Hosted by Bob (Jue) Guo — CS PhD candidate and adjunct faculty at the University at Buffalo, teaching deep learning, machine learning, and AI.