anlak storytelling
All kinds
Document kinds
Loading document kinds...

Primary corpus lens for browsing and insights. Selection lives in the URL.

All formats
Orientation
Loading formats...

Landscape, portrait, square, or unknown document geometry.

Unlock
OpenAI · conference-presentation

OpenAI | Product Presentation Deck | 52 slides

52 pages · 4 arc beats · 1 loops

Conference-Presentation Download JSON
Unlock

OpenAI | Product Presentation Deck | 52 slides

OpenAI · 2018-11
arc beats above · slides in the middle · loops below · scroll → 1 LOOPS
SETUP TENSION ANALYSIS EVIDENCE RESOLUTION APPENDIX
HOVER FOR DETAILS · CLICK A SLIDE FOR FULLSCREEN · STEP 2
— No image bound to this slide. Wire up imgproxy to display the rendered slide JPEG.
—

Deck intelligence map

4
coverage by narrative range · generated from this deck JSON
Narrative range 19 total
Metadata
Components
Metrics
Tools
Frameworks
Beats
Loops
What Is 4 slides 100% 4/4 slides 100% 4/4 slides · 9 hits
— 0/4 slides
50% 2/4 slides
— 0/4 slides
100% 4/4 slides
— 0/4 slides
Problem (Identify pain) 4 slides 100% 4/4 slides 100% 4/4 slides · 8 hits
— 0/4 slides
25% 1/4 slides
— 0/4 slides
100% 4/4 slides
— 0/4 slides
Solution (Provide relief) 6 slides 100% 6/6 slides 100% 6/6 slides · 17 hits
— 0/6 slides
33.3% 2/6 slides 16.7% 1/6 slides 100% 6/6 slides
— 0/6 slides
Why 5 slides 100% 5/5 slides 100% 5/5 slides · 12 hits
— 0/5 slides
— 0/5 slides
— 0/5 slides
100% 5/5 slides
— 0/5 slides

Slide inventory

52
every slide · same image gating as the playbook
01
Slide 1
front_matter
Open slide detailBeat · What Is
02
establish_context
Open slide detailBeat · What Is
03
transition
Open slide detailBeat · What Is
04
front_matter
Open slide detailBeat · What Is
05
transition
Open slide detail
06
frame_problem
Open slide detailBeat · Problem (Identify pain)
07
establish_context
Open slide detailBeat · Problem (Identify pain)
08
The slide serves as a visual example of the environment in which the OpenAI Five AI agent operates.illustrate_case
Open slide detailBeat · Problem (Identify pain)
09
present_framework
Open slide detailBeat · Problem (Identify pain)
10
frame_situation
Open slide detailBeat · Solution (Provide relief)
11
The chart tracks performance milestones from May to August 2018, showing a steady increase in MMR.analyze_data
Open slide detailBeat · Solution (Provide relief)
12
front_matter
Open slide detailBeat · Solution (Provide relief)
13
The slide shows a robotic hand manipulating a cube with letters on it, labeled 'GOAL 1' with a target cube configuration.illustrate_case
Open slide detailBeat · Solution (Provide relief)
14
The slide serves as a visual example of robotic dexterity with varied objects.illustrate_case
Open slide detailBeat · Solution (Provide relief)
15
Uses a split-screen layout to contrast physical hardware with digital simulation.present_framework
Open slide detailBeat · Solution (Provide relief)
16
The slide demonstrates the concept of training AI models in simulation by varying environmental parameters to improve real-world robustness.present_framework
Open slide detail
17
The diagram depicts a neural network architecture involving convolutional layers (CONV) and a recurrent layer (LSTM) for robotic manipulation.present_framework
Open slide detail
18
transition
Open slide detail
19
present_framework
Open slide detail
20
RND likely refers to Random Network Distillation, a reinforcement learning technique.illustrate_case
Open slide detail
21
The slide uses a retro video game screenshot as a visual metaphor or reference point.other
Open slide detail
22
The chart shows a learning curve plateauing around 8k return.analyze_data
Open slide detail
23
The chart shows a learning curve for a reinforcement learning agent, with a red dot indicating a specific performance milestone.analyze_data
Open slide detail
24
The chart tracks various reinforcement learning algorithms (SARSA, DQN, A3C, RND, etc.) from 2013 to 2019.analyze_data
Open slide detail
25
filler
Open slide detail
26
The slide illustrates reinforcement learning agent behavior, specifically intrinsic motivation/curiosity.analyze_data
Open slide detail
27
The chart shows a single red dot at x=2500, y=0.2, and a vertical line spike near x=2800.analyze_data
Open slide detail
28
transition
Open slide detailBeat · Why
29
establish_context
Open slide detailBeat · Why
30
summarize
Open slide detailBeat · Why
31
summarize
Open slide detailBeat · Why
32
frame_situation
Open slide detailBeat · Why
33
transition
Open slide detail
34
The diagram is a standard representation of a multi-layer perceptron (MLP).establish_context
Open slide detail
35
The slide highlights the limitations of older computer vision techniques (HOG features) by showing a false positive detection.illustrate_case
Open slide detail
36
The chart illustrates the rapid improvement in computer vision models over a 5-year period.analyze_data
Open slide detail
37
The diagram shows a standard RNN/LSTM encoder-decoder architecture.analyze_data
Open slide detail
38
The slide shows two grids of generated images, likely comparing different datasets or model outputs from the original Goodfellow et al. GAN paper.illustrate_case
Open slide detail
39
The slide displays two grids of AI-generated images, likely from the DCGAN paper by Radford et al. (2015).illustrate_case
Open slide detail
40
The slide displays a single AI-generated portrait as a case study of GAN progress.illustrate_case
Open slide detail
41
The slide showcases the state of GAN-based image generation as of 2018.illustrate_case
Open slide detail
42
The slide highlights the 2013 Mnih et al. paper on DQN.illustrate_case
Open slide detail
43
The slide features a visual demonstration of a humanoid agent learning to walk/balance in a 3D environment.illustrate_case
Open slide detail
44
The slide features a screenshot from the 2016 AlphaGo match, highlighting the application of reinforcement learning in complex strategy games.illustrate_case
Open slide detail
45
The slide highlights the computational scale (+100,000 CPU cores, +1000 GPUs) used for OpenAI Five.illustrate_case
Open slide detail
46
transition
Open slide detail
47
The slide uses a simple neural network diagram to illustrate the concept of parallelization.establish_context
Open slide detail
48
The chart highlights the massive jump in compute requirements starting around 2017 with AlphaGoZero.analyze_data
Open slide detail
49
The slide uses a simple list format to highlight open research problems.diagnose_problem
Open slide detailLoop · Cost Of Inaction
50
transition
Open slide detailLoop · Cost Of Inaction
51
frame_problem
Open slide detailLoop · Cost Of Inaction
52
summarize
Open slide detail