AlphaGo, OpenAI’s Dota Bot and What a Result Means
RAISE GENIUS · PARENTS & SCHOOLS
AlphaGo, OpenAI’s Dota Bot and What a Result Means
Learning from game playing AI requires knowing which system and rules were involved. This archive connects AlphaGo with OpenAI’s 2017 Dota 2 demonstration.
Contact the teamOn this page
Learning from game playing AI is not a humanoid demonstration
The Thai title refers to intelligent robots imitating humans, while the body mainly concerns software playing games. A game-playing AI does not need a human-shaped robot body.
The source introduces AI through several disciplines and an interest in behaviour associated with intelligence. Performance in one setting should not be taken as evidence of consciousness, human emotions or a system that can do every job.
AlphaGo and a particular Go milestone
The article refers to Go matches with Ke Jie. The relevant 2017 milestone is different from the 2016 match with Lee Sedol, which AlphaGo won four games to one.
The source’s statement that AI won every match should therefore not become a universal claim. Versions also differ: the earlier AlphaGo and later AlphaGo Zero did not use exactly the same learning setup.
OpenAI’s 2017 one-versus-one Dota bot
OpenAI’s contemporary report describes a bot learning through self-play and a one-versus-one demonstration against Dendi at The International on 11 August 2017. The archive incorrectly calls OpenAI a Tesla company; that is corrected here.
The demonstration was not the complete five-versus-five game or evidence that the bot could play every hero, rule set or real-world task. The official report describes its defined setting and the work still ahead.
Use the result as a question, not a prediction of everything
The archive imagines future robots as companions and encourages technology learning. That is speculation rather than a verified ability of the demonstrated systems or a timetable for replacing people.
Learning from game playing AI can begin by asking what the system observed, how success was defined and what the test left out. The original video is retained as an optional link without an automatically loaded player.
Original video archive
Choose a link to open the original video on YouTube. No external player, automatic playback or YouTube thumbnail is loaded here. The clips are retained from the published archive; their language, availability and content belong to the original upload.
Frequently asked questions
Was the 2017 Dota demonstration the full five-versus-five game?
No. The contemporary report concerns a defined one-versus-one setting.
Is OpenAI a Tesla company?
No. The archive’s attribution is incorrect.
Does winning a game demonstrate human emotions?
No. A game result does not establish consciousness or human feelings.
Contact and related pages
Tell the team about the learner or organisation, relevant experience and the intended learning purpose. Confirm the current programme, equipment, availability, schedule and fees before booking.
Contact Raise Genius School service overview
- Teacher Num on Work, Technology and Changing Possibilities
- Observe, Test and Adapt an Idea with Support
- Research, Social Robots, Electronics and Useful Machines
- Cleaning, Cooking and Folding as Design Questions
- Bitcoin, AlphaGo and a Child’s First Design Ideas
- Technology Interest, School Workload and Playful Learning
- Keep the Activities, Clarify the Brain-Side Claim
- Collaboration across Thai Universities and Institutes
- A Small Robot, Coloured Paths and Planning
Course-selection guide · Robotics portfolio projects · Competition preparation · Team advice and coaching · Course guide · General FAQ
Related services: School robotics camps · STEM activity stations
Related learning: Raise Maker Arduino · micro:bit robotics · IOTBot & AI · Portfolio preparation · Camp proposal · Accompanying teachers
Related reference: OpenAI: Dota 2, 11 August 2017 · Google DeepMind: AlphaGo’s next move, May 2017 · Google DeepMind: AlphaGo
