I might be slightly biased since my current focus for the past few years has been on Artificial Intelligence, but it seems like, in this past year, there were many more AI talks at major gaming conferences than in the past (most were restricted to AI summits but now there seemed to be an AI talk pretty much every hour).
China is already beating us on any kind of ML (Machine Learning) tech advancements, so we’ll be playing catch-up for the foreseeable future.
That said, some of the particular systems I’ve personally been working on over the past few years offer really innovative ways of handling ML (such as doing designer-authored behavior trees/branches/group behaviors and/or using ML for selection/balancing/difficulty) and I’d love to delve deeper into that realm and make some similar use-cases for future products.
However, Tencent/Wukong AI is probably the craziest/most advanced AI, driven by deep learning/imitation learning I’ve seen, and let me explain why…but first some context.
What Is AI? (a brief overview for those not in the field)
“Artificial intelligence (AI) plays an increasingly central role in this transformation. In recent years, AI has come roaring out of research laboratories to become ubiquitous and ambient in our personal lives, so much so that many consumers do not realize they use products and applications that contain AI on a daily basis.
AI stands to greatly benefit all industries, achieving adoption leaps from consumer segments to enterprises and onward to the industrial sector. Technological progress in the fields of big data, algorithmic development, connectivity, cloud computing and processing power have made the performance, accessibility, and costs of AI more favorable than ever before. Just as the relational database found its way into core business operations around the world – providing better ways to store, retrieve, and organize information – AI is now following a similar path. It is becoming an integral part of every future software system and soon we will no longer need to call it out as AI.“
“In the ‘Age of AI’, where ‘data’ is the new ‘oil’, China is the new Saudi Arabia”
In a (somewhat) recent special by PBS, they touch on the rise of AI in the Eastern markets, in particular they discuss Google China and how in the near future, AI is going to be more important on the world-stage that oil.
The key take-away here is that, even though this video only goes up to the same time as the graph above (about 4 years ago when Alpha GO beat all but one single female professional GO player), it has been known and identified that AI is going to play a seminal role going forward not just for games but in the world economy.
Titan A.T. (After Tencent)
It’s no surprise to anyone working in the game development field that TenCent is a GIANT (some would say Titan) of the gaming industry. Originally founded in 1998 to develop WeChat (the premier chat platform in China), the name “Tencent” is based on its Chinese name Tengxun (腾讯), which incorporates part of Pony Ma’s Chinese name (Ma Huateng; 马化「腾」) and literally means “galloping fast information”.Source
In 2003, they branched out tackling the gaming market and have since become the “world’s largest video game vendor”. Since then, they have come to wholly or partly own some of the biggest gaming franchises in the world. Riot, Epic, Discord, Activision, Hi-Rez (shout out to my buddies from Full Sail). They now publish Honor of Kings (a rip-off of League, which was a rip-off of DOTA-All Stars, which was a rip-off of DOTA, which was in turn a rip-off of Aeon of Strife – if you don’t know your MOBA history, I suggest you go check it out), which according to their publicly announced earnings grossed $216M in a single month (Feb 2021) and currently earns close to $2.6B annually (adjusted for exchange rates).
That said, games are only one part of the conglomerate. Their services include social networks, music, web portals, e-commerce, mobile games, internet services, payment systems, smartphones, and multiplayer online games.
What is ‘Wukong AI’ and why is it so important?
So, anyone who is a total nerd like me keeps up with the latest AI publications and white papers. I started doing it when studying astrophysics and have kept with the habit over the years. I go to all the conferences (when possible) as well as get all the publications. Normally, it costs a TON of money, or you have to be a student to access the ‘Scientific Journals’, but recently there have been a lot of advancements in sharing of knowledge (more on that later in this article as it relates to China).
I was going through Research Gate (one such source making Scientific Journals more accessible), which regularly hosts all of the papers published by Cornell University (one of the more publishing-focused college institutions in the USA; and coincidentally, one of our partners when I was working on the Mars Rovers), and came across a paper entitled: Mastering Complex Control in MOBA Games with Deep Reinforcement Learning
Earlier in the year, I had been doing research for a side-project that involves machine learning and had come across “A Survey of Planning and Learning In Games” that provides an excellent overview, taking everything from the very first graph in this article and mapping it out over a period of time as it relates to games as well as covering a vast majority of other content/subjects related to the various applications that can be made for ML in games.
It touches on several critical points in the timeline including:
- 1997 when Deep Blue beat Kasparov,
- 2011 when Watson beat the top Jeopardy players
- and more…
(to be continued in Part 2)…
Ben Retan 