The overall illustration of this thread!
Note: Illustration generated using NotebookLM & original lecture from Yann Dubois.
#AgenticAIMOOC#AgentBeats@BerkeleyRDI
Over the last 3 months, I took an AI Agentic course at UC Berkeley.
As part of the final assignment, I'd like to summarise the key concepts taught during the lecture from Yann Dubois on LLM Agents Overview. Here is π§΅
#AgenticAIMOOC#AgentBeats@BerkeleyRDI
12/ GPU Bottleneck: Modern training isn't limited by calculation speed, but by memory and communication.
Techniques like Mixed Precision (bf16) and FlashAttention (using tiling and fusion) are used to keep the processor "fed" with data efficiently. #AgenticAIMOOC#AgentBeats
I've never felt this much behind as a programmer. The profession is being dramatically refactored as the bits contributed by the programmer are increasingly sparse and between. I have a sense that I could be 10X more powerful if I just properly string together what has become available over the last ~year and a failure to claim the boost feels decidedly like skill issue. There's a new programmable layer of abstraction to master (in addition to the usual layers below) involving agents, subagents, their prompts, contexts, memory, modes, permissions, tools, plugins, skills, hooks, MCP, LSP, slash commands, workflows, IDE integrations, and a need to build an all-encompassing mental model for strengths and pitfalls of fundamentally stochastic, fallible, unintelligible and changing entities suddenly intermingled with what used to be good old fashioned engineering. Clearly some powerful alien tool was handed around except it comes with no manual and everyone has to figure out how to hold it and operate it, while the resulting magnitude 9 earthquake is rocking the profession. Roll up your sleeves to not fall behind.
@gregisenberg Agree - this is happening with all AI application companies.
Recently Elena Verna shared the same view during Lennyβs podcast. There is no single PMF rather a series of PMF with same core group of users.