Talks
Research Seminar June 15, 2023 · 119 slides

Everything Everywhere All at Once

Comprehensive overview of AI research directions — NLP, ESG, sustainability, and responsible AI integration.

AI NLP ESG Sustainability

Overview

This talk, subtitled Infinite Possibilities: AI’s Journey Towards Understanding Everything, Everywhere, All at Once, takes its name and spirit from the film — a meditation on complexity, interconnectivity, and the possibility of holding contradictory truths simultaneously. AI research is at exactly such an inflection point.

The presentation opens by asking what intelligence actually is. Using IBM’s Watson as a foil, it draws a distinction between information retrieval and genuine understanding. Watson’s Jeopardy! victory was a triumph of fast pattern-matching over a structured corpus — impressive, but fundamentally different from the world-modelling that underlies human reasoning. The central thesis is that the next frontier is not bigger models but better world models: internal representations that let AI systems interpret context, not just retrieve tokens.

Key Themes

From Turing to Transformers. The talk traces a conceptual line from Alan Turing’s Imitation Game through the Neural Turing Machine, attention mechanisms, and the Transformer architecture. Each step is framed as an attempt to move from rote imitation toward genuine comprehension.

Large Language Models and Their Limits. GPT-4 and its contemporaries generate impressively coherent text, yet they lack grounded world models. The presentation examines hallucination, knowledge graphs, and retrieval-augmented generation as partial solutions to this gap.

Cutting-Edge Tools. The second half surveys the 2023 toolkit: LangChain, LlamaIndex, Segment Anything Model (SAM), DINOv2, and ImageBind — situating each within the broader arc from perception to reasoning.

ESG and Responsible AI. The talk closes by connecting AI capability to societal responsibility, arguing that understanding everything carries an obligation to act well — linking the technical themes to the ESG and sustainability research that runs throughout this body of work.

The 119-slide deck is dense by design: the goal is to convey the sheer breadth of the AI landscape as it stood in mid-2023, and to argue that only researchers comfortable moving across all of it will be able to shape what comes next.

Slides

119 slides
Slide 1
Slide 2
Slide 3
Slide 4
Slide 5
Slide 6
Slide 7
Slide 8
Slide 9
Slide 10
Slide 11
Slide 12
Slide 13
Slide 14
Slide 15
Slide 16
Slide 17
Slide 18
Slide 19
Slide 20
Slide 21
Slide 22
Slide 23
Slide 24
Slide 25
Slide 26
Slide 27
Slide 28
Slide 29
Slide 30
Slide 31
Slide 32
Slide 33
Slide 34
Slide 35
Slide 36
Slide 37
Slide 38
Slide 39
Slide 40
Slide 41
Slide 42
Slide 43
Slide 44
Slide 45
Slide 46
Slide 47
Slide 48
Slide 49
Slide 50
Slide 51
Slide 52
Slide 53
Slide 54
Slide 55
Slide 56
Slide 57
Slide 58
Slide 59
Slide 60
Slide 61
Slide 62
Slide 63
Slide 64
Slide 65
Slide 66
Slide 67
Slide 68
Slide 69
Slide 70
Slide 71
Slide 72
Slide 73
Slide 74
Slide 75
Slide 76
Slide 77
Slide 78
Slide 79
Slide 80
Slide 81
Slide 82
Slide 83
Slide 84
Slide 85
Slide 86
Slide 87
Slide 88
Slide 89
Slide 90
Slide 91
Slide 92
Slide 93
Slide 94
Slide 95
Slide 96
Slide 97
Slide 98
Slide 99
Slide 100
Slide 101
Slide 102
Slide 103
Slide 104
Slide 105
Slide 106
Slide 107
Slide 108
Slide 109
Slide 110
Slide 111
Slide 112
Slide 113
Slide 114
Slide 115
Slide 116
Slide 117
Slide 118
Slide 119