Artificial Gamer Documentary Screening March 2022

AI Summary

A documentary screening, "Artificial Gamer," showcases OpenAI's efforts to develop an AI bot capable of defeating Dota 2 world champions within a year. The film explores the team's push for machine learning advancements, including self-play and reinforcement learning, which enable the AI to learn strategies and react faster than humans.

A passionate team of Silicon Valley engineers at OpenAI is in a race to develop an artificial intelligence bot capable of defeating the world champions of one of the most lucrative esports – Dota 2. And they’ve got one year to do it. The outcome could alter the way we think about the frontiers of machine learning and of humanity’s relationship to AI.

For more than 80 years humans and machines have competed in chess, Go, poker, Jeopardy! The latest battlefields are hugely popular multiplayer online games – like Dota 2 – that are worldwide sensations. In tihs in-person screening of Artificial Gamer, it talked about OpenAI’s efforts to build the most powerful esports bot. It begins at the 2017 annual eSports world championship, also known as “The International.” OpenAI announces that an AI agent will play against some of the world’s best DOTA 2 players. Dota 2 is a multilayered experience involving bluffing, coordinated five-player strategy, and epic psych-outs — a level of gaming considered too sophisticated for the traditional reach of computers. As OpenAI’s software not only successfully matches the world’s top DOTA 2 players, but wins by a landslide in one-on-one competition, the creators behind the AI seek to step up their game and create a software that can successfully compete and win against five players in a five-on-five match.

1. Overall Rating (0–10) — 6.0
This photograph captures the atmosphere of a panel discussion on a modern stage, with a strong sense of event energy and thematic focus. The bold blue lighting and central screen create a cohesive, tech-forward mood, though the composition feels slightly unbalanced due to the wide-angle perspective. While the image effectively conveys the setting and subject matter, it lacks the visual precision to feel fully immersive or artistically refined.

2. Composition (0–10) — 5.5
The wide-angle framing includes too much of the surrounding space, diluting focus on the panelists. The stage is centered, but the audience’s heads in the foreground disrupt the clean view of the speakers.

3. Lighting (0–10) — 7.0
The deep blue wash from overhead fixtures creates a dramatic, immersive atmosphere that complements the tech-themed presentation. The screen is well-lit and readable, though some shadows fall unevenly on the stage floor.

4. Color & Tone (0–10) — 6.5
The dominant cool blue palette enhances the futuristic theme, but the overall tone is somewhat flat due to limited contrast. The red chairs provide a striking visual anchor, but the neutral tones of the stage and audience mute the vibrancy.

5. Creativity (0–10) — 6.0
The image effectively documents a live event with clear thematic intent, but it relies on standard event photography conventions rather than pushing artistic boundaries. The framing and lighting choices are functional, not inventive.

6. Technical Quality (0–10) — 7.0
The image is sharp and free of major technical flaws, with good focus on the stage and screen. The wide-angle lens introduces some distortion, but it doesn’t significantly detract from clarity.

7. Emotional Impact (0–10) — 5.5
The photograph conveys a sense of intellectual engagement and modernity, but the viewer remains at a distance due to the lack of intimate framing or emotional focus on the participants.

The documentary went on interviewing key players in the field, displaying fun and exciting sequences of gameplay that effectively change the pace of the film, and delivering a solid story arc balancing the viewpoints of the OpenAI creators as well as the DOTA 2 community. I enjoyed learning how the AI developers push the boundaries of machine learning, extracting and repurposing algorithmic learning to take advantage of processing power that can cram 180 years’ worth of Dota play every day, learning via self-play. It trains using a scaled-up version of Proximal Policy Optimization running on 256 GPUs and 128,000 CPU cores — a larger-scale version of the system built to play the much-simpler solo variant of the game.

Using a separate LSTM for each hero and no human data, it learns recognizable strategies. This indicates that reinforcement learning can yield long-term planning with large but achievable scale — without fundamental advances, contrary to our own expectations upon starting the project. OpenAI Five is given access to the same information as humans, but instantly sees data like positions, healths, and item inventories that humans have to check manually. Our method isn’t fundamentally tied to observing state, but just rendering pixels from the game would require thousands of GPUs. OpenAI Five averages around 150-170 actions per minute (and has a theoretical maximum of 450 due to observing every 4th frame). Frame-perfect timing, while possible for skilled players, is trivial for OpenAI Five. OpenAI Five has an average reaction time of 80ms, which is faster than humans. More information can be found here.

1. Overall Rating (0–10) — 6.0
This photograph captures a promotional flyer for a tech-focused event, conveying a professional and informative tone. The layout is clean and structured, with clear typography and purposeful visual hierarchy, though the image’s candid nature—held in hand with a dark background—undermines its potential aesthetic impact. While the content is compelling, the execution feels more like documentation than art, lacking the deliberate staging or emotional resonance to elevate it beyond a functional record.

2. Composition (0–10) — 6.0
The flyer is centered and angled slightly, with the hand and background framing it unevenly. The composition prioritizes legibility over visual harmony, with the text and portraits arranged logically but not artistically. The depth created by the hand and background adds context but distracts from the flyer’s intended prominence.

3. Lighting (0–10) — 5.0
The lighting is flat and ambient, likely from an indoor source, casting soft shadows and creating uneven exposure across the flyer. The darker background contrasts with the white paper, but the lack of directional or mood lighting results in a neutral, unemotional atmosphere.

4. Color & Tone (0–10) — 6.0
The palette is predominantly monochromatic—black, white, and gray—with the red accent in Marguerite Gong Hancock’s photo providing a subtle pop of color. The tone is restrained and corporate, aligning with the event’s technical subject matter, though the lack of vibrancy limits visual engagement.

5. Creativity (0–10) — 5.0
The creativity lies in the design of the flyer itself—functional, modern, and informative—but the photograph’s purpose is observational rather than expressive. The image captures the event’s identity well, but offers little originality in its framing or interpretation.

6. Technical Quality (0–10) — 7.0
The focus is sharp on the flyer, with clear legibility of text and portraits. The image is free of major technical flaws, though slight motion blur and uneven lighting reduce its overall polish.

7. Emotional Impact (0–10) — 5.5
The image evokes a sense of anticipation and intellectual curiosity, particularly for those familiar with AI and esports. However, the emotional resonance is muted by the photo’s documentary nature and lack of personal connection or narrative depth.

Following the screening, there is a panel discussion and Q&A featuring Open AI team’s David Farhi and Susan Zhang. Some backgroudn about David Farhi who is a research lead at OpenAI, where he led the DotA 2 project during its seminal achievement of superhuman performance in a top competitive e-sport. He studies reinforcement learning in complex environments and is especially interested in environments with unbounded emergent behavioral complexity. He completed his PhD at Harvard in theoretical physics and undergraduate degree at MIT.

1. Overall Rating (0–10) — 6.8
This photograph captures a futuristic vehicle in a museum-like setting, where the interplay of technology and human-scale design creates a sense of quiet anticipation. The open door invites curiosity, while the clean, bright environment underscores the vehicle’s innovation. While the composition is clear and informative, it lacks the dramatic tension or emotional depth that would elevate it beyond a straightforward documentation.

2. Composition (0–10) — 6.0
The car is framed slightly off-center with the sign partially obstructing the left side, creating a slight imbalance. The open door draws attention, but the background elements—windows, trees, and distant furniture—distract from the vehicle’s sleek form.

3. Lighting (0–10) — 7.0
Natural light from the large windows illuminates the scene evenly, highlighting the car’s smooth curves and reflective surfaces. The soft, diffused lighting enhances the futuristic aesthetic without creating harsh shadows.

4. Color & Tone (0–10) — 6.5
The neutral palette—whites, grays, and blacks—complements the vehicle’s minimalist design. However, the lack of color vibrancy gives the image a sterile feel, slightly dampening the sense of technological wonder.

5. Creativity (0–10) — 6.5
The image successfully conveys the concept of autonomous transportation, but it remains largely literal and documentary in nature. A more dynamic angle or narrative framing could have added conceptual depth.

6. Technical Quality (0–10) — 7.5
The image is sharp and well-focused, with clean details on the car’s body and interior. The depth of field is appropriate, keeping both the vehicle and the sign in clear view.

7. Emotional Impact (0–10) — 5.5
While the vehicle evokes a sense of progress and innovation, the image’s clinical presentation keeps the viewer at a distance. The emotional resonance is muted, leaning more toward intellectual curiosity than personal connection.

In the front lobby area of the Computer History Museum, there is also a Waymo (part of Google) self drive car, called Firefly. Google initially designed the Firefly back in 2013. From the start, the car was intended to be an experiment that would allow engineers to explore different ideas about how autonomous vehicles should work or be configured, they said. Although it was seen all over the place in cities like Mountain View while Google was testing it, the Firefly was never intended to be a production vehicle. Along the way, the Firefly racked up some impressive achievements for the development of autonomous vehicles, including million of miles driven and the first completely autonomous trip. Visitors are allowed to get in the Waymo and try out the ride the auto seldrive car. Pretty cool!