"Trauma Team mbers, this is just one of the professions the protagonist in Cyberpunk 2077 can take on. What you just saw was rely one small mission from that profession. Of course, the mission was clearly arranged as part of a staged performance, but creating such a mission is actually simpler than you might think."
In traditional ga combat level design, there are often a variety of complex variables, obstacles, and other intricate elents.
Designing a compelling level that players can get hooked on is no easy feat.
Even in turn-based combat, which is considered a lower-intensity mode, there are still very intricate calculations involved in balancing the gaplay.
In the past, all of this required massive manpower and programrs working overti.
However, with the help of AI today, most of the complicated work can be entirely handled by artificial intelligence.
Developers now simply need to wait while the AI generates a vast number of levels through large-scale computation and simulations. Afterward, the developers manually review the generated content, pick out the most suitable levels, or even select all of them if they et the quality standard.
So levels, thanks to repeated AI simulations, are so well-made that they can be integrated directly into the ga's main story—these are the high-quality gems worth using without hesitation.
Of course, such miracles can't always be expected. It's unrealistic to hope AI will continuously produce top-tier content on its own.
But creating a combat scenario like the one earlier is relatively easy for AI.
First, AI-generated thugs are arranged and ard. Then, several ambushed Trauma Team mbers are placed outside the building. Using big data, AI runs countless simulations to determine the most visually impactful combat setup.
Afterward, the team manually selected the battle scenario you just saw.
The first wave of the Trauma Team gets wiped out, highlighting the brutality of the city, followed by more reinforcents sacrificing themselves to rescue two rich clients.
This structure was already well demonstrated in earlier gas like Assassin's Creed and GTA: The Chapter of Liberty City.
However, those two gas were more like experintal fields. Once the thod was perfected there, it was applied directly to Cyberpunk 2077, speeding up developnt significantly.
Originally, Cyberpunk 2077 was expected to take at least another year or two to fully complete.
And even then, it likely wouldn't have reached the current level of quality.
"Gastar Electronic Entertainnt has truly mastered AI computation to an art form."
Even Hayakawa Ueto was watching the livestream.
Their own exhibit section hadn't opened to the public yet, giving him ti to watch Yuxing Electronic Entertainnt's presentation with others.
This ti, Gastar was the first to reveal all their cards.
And their opening move was a royal flush—ga developnt technology that other companies couldn't even dream of matching.
"This is really hard to keep up with."
"Don't lose hope—we expected this. That's why we made such thorough preparations, right?"
As company president, Hayakawa Ueto cheered on his team, calming their nerves.
"He's right. We should have seen this coming."
"Let's just quietly watch the rest of their presentation. It's probably not even over yet."
Despite trying to stay composed, Hayakawa couldn't help feeling conflicted.
Gastar Electronic Entertainnt's technological gap with other ga developers was massive.
To put it in perspective: in AI application, Gastar is operating at racecar level, while everyone else is still driving family sedans.
The speed gap isn't sothing that can be closed quickly.
The Gastar presentation continued.
The Trauma Team was just one of the protagonist's identities.
In the original ga, players could choose from three backgrounds: Corpo (company dog), Street Kid, or Nomad (scavenger).
These three identities corresponded to the most distinct social classes in Night City.
Back when these roles were announced, Takayuki hoped the ga would feature three distinct main storylines that perhaps led to the sa ending in different ways.
But after actually playing the ga, he was disappointed to find that the three identities only affected the tutorial. The main story was exactly the sa for all three.
Clearly, CD Projekt Red had either cut corners or lacked the ti and budget to expand the content.
Sotis you have grand ideas, but reality in the form of limited resources crushes them.
But not in this world.
In this world, Takayuki went further and added even more detail.
He then showcased several of the protagonist's other identities in Cyberpunk 2077:
A violent gang mber (as in the earlier demo)
A company dog lost in luxury and excess
A useless person addicted to the VR world of "Oasis"
Soone spiritually trapped forever in the simulated world of "Braindance"
Each identity ca with a completely different main storyline.
Just these multiple storylines and performances alone amounted to tens of millions of lines of dialogue.
The voice actors for the male and female protagonists nearly suffered vocal injuries from the constant shouting and emotional delivery.
As for all the other characters, the situation wasn't any better.
This made Takayuki start to seriously look forward to AI voice acting in the future.
Once that becos available—and emotionally expressive—it will enable massive-scale voice production without needing breaks.
Artificial intelligence, automation, and robotics were originally invented to relieve humans from tedious labor, not to replace them.
Can you imagine a future video ga with tens of millions of lines of dialogue?
If done manually, how many years would it take?
Ten? Twenty? Fifty?
With AI, it's no longer a problem.
Reviews
All reviews (0)