candy crush has been using artificial intelligence in its development processes for years. For King, the company responsible for the game, neural networks and machine learning They are not something new. In fact, they have been using artificial intelligence since before 2022, when King acquired the Swedish company Peltarion, specialized in AI.
The use of artificial intelligence is controversial. They help process immense amounts of data, predict behavioral patterns, and can be used to make our lives easier. But they also serve to measure to the millimeter each step we take in games and applications, the management of big data it can be used in measures against users and then there is generative AI and all the problems that come with it.
Read also
King’s studios in Stockholm. Candy Crush 10th Anniversary
At King they have their own Artificial Intelligence laboratory and use the machine learning in several of its development processes. The specialized medium Games Industry has interviewed the director of this laboratory, Sahar Asadi, who has told how they use AI to create levels of candy crush.
Checking levels is 50% faster
The main use of AI in the development of candy crushAccording to Sahar Asadi, it is in the correction and iteration of the levels. “The bot playtesting that we have developed gives designers, before publishing a level, a greater understanding of the gaming experience,” says Asadi. This is a game testing bot that evaluates levels to see what errors they have and suggests possible solutions.
Image of Cany Crush screen
Sahar Asadi explains that the bot is designed to behave like a human would. “It’s not designed to win, it’s built to mimic human behavior,” he says. To do this, engineers have trained the AI with thousands of examples from real users and rely on data generated by gamers. candy crush daily.
In addition, this artificial intelligence is also capable of suggesting changes and modifications to levels so that they adapt to the difficulty and the desired experience. “The fun part for designers is creating the levels. The mundane part is iterating, playing the level, seeing what it is like, and if you’re not happy, going back and tweaking it. This manual work is mundane. The bot playtesting It helps reduce the time spent on these corrections and allows more time to be spent on the creative part,” says Asadi.
Image from the video game ‘Candy Crush’
In terms of numbers, the director of King’s AI laboratory states that the testing bot has helped reduce manual corrections made to levels by 95% and that this translates into 50% less time spent correcting the levels.
Dangers and doubts about AI
A few years ago, these review and correction processes were carried out by less experienced designers. It was a good way to get familiar with the game and the level design. AI takes these jobs out of the equation. Yes, it helps increase productivity, but it eliminates jobs and makes it difficult for people with little experience to enter the industry.
Image from the video game ‘Candy Crush’
“We need designers, we see this tool as a help for them. In the end, it is the designers who know what is fun and what gaming experience they want to get,” says Asadi when asked about the risks of AI. “If you go back in time, designers used paper and pencil to work, then they moved on to Photoshop and new tools. “I see AI as another new tool that allows them to focus on what they are really good at,” he says.
“The way of working will change,” Asadi continues, “the skills you will need in your day-to-day work will change. Thanks to machine learning many engineers no longer need to program. When I interview candidates, I no longer base myself on the same criteria as a few years ago.”
‘Candy Crush’ and the future of AI
Image from the video game ‘Candy Crush’
In 2022, candy crush It had more than 210 million users worldwide. With more than 5 billion downloads, it will also reach the figure of 20 billion dollars in profits in 2022. candy crush It is one of the largest, most profitable and most played video games in the world.
From King they assure that they will continue implementing AI in their star game. Sahar Asadi points out that, in the not too distant future, candy crush could be able to adapt the difficulty and duration of the levels to the context of each player, thus personalizing the experience.
A YOUNG GIRL PLAYING CANDY CRUSH WITH THE MOBILE ON THE TRAIN
On the other hand, it is easy to foresee that with the increase in productivity associated with AI, jobs in the industry will become increasingly specialized, while workforces are reduced and more mass layoffs are announced.
Read also
Video games
#Candy #Crush #improve #levels
