slide 6
slide 6b — morning
slide 6c — afternoon
slide 7a — speaker bio (verbatim from the IMAM 2.0 AMJA workshop deck)
slide 7b — speaker bio (refreshed from the vric26 bio)
slide 25c — red pill / blue pill (reinserted from the IMAM 2.0 AMJA workshop deck, delivered a year ago at this venue)
Speaker notes (verbatim from the IMAM 2.0 deck): Blue pill: Red pill: - Trained on its own data, model collapse, we will lose expertise and models will get worse. Can we recover? - Accuracy is just not high enough for use in many tasks, may not get there - Easily co-opted by governments, billionaires to surveil, manipulate public opinion, elections, etc. - Brain rot and loss of expertise - MASSIVE electricity usage. UK gov't told citizens to delete their photos from the cloud to lower cloud costs - Data haves and have-nots. Concentration of power and wealth This doesn't take away from any of the other points which is that we absolutely have to take ownership, leverage it, be on top of this revolution. If we don't, we will be leveraged.
slide 8
slide 11
slide 16
slide 18a
slide 18b
slide 18b2 — benchmarks
slide 31
slide 33
slide 36a
slide 36b
slide 36c
slide 38
slide 40
slide 43
slide 45
slide 47
slide 48
slide 49
slide 49b — verbatim from gmc26 "Where AI Video Is Today"
slide 51
slide 55
slide 57
slide 59
slide 61
slide 65a
slide 65b
slide 66
slide 70a
slide 70b
slide 72
slide 73
slide 75
slide 81
slide 83
slide 84
slide 86
slide 91
slide 93
slide 95
slide 99
slide 97
slide 100
slide 100b
slide 101a
slide 101b
slide 103
slide 105a
slide 105b
slide 105c
slide 107b
slide 109a
slide 109b
slide 111
slide 113
slide 115
slide 116a
slide 116b
slide 118a
slide 118b
slide 118c
slide 124
slide 126
slide 128a
slide 128b
slide 130
slide 132a
slide 132b
slide 134
slide 136
slide 138
slide 140
slide 142a
slide 142b
slide 144a
slide 144b
slide 146
slide 150
slide 152
slide 154
slide 154a
slide 154b
slide 154c
slide 154d
slide 154e
slide 154f
slide 155 — the prompt for general AIs (from the JaleesBench companion-prompt article)
slide 156
slide 158
slide 161
slide 164
slide 166
slide 168
mit26 Part 2 — Ansari case study
slide 211a
slide 211b
slide 214
slide 216a
slide 216b
slide 219
slide 219b — callback to "Teach a daee to fish" (intro)
slide 223
slide 224 — companion prompt callback (s.iaser.ai/prompt)
slide 224b — the Claude ladder
slide 224c — Gemini Spark
slide 231a
slide 225
slide 227
slide 233
slide 234