This game is awesome

The administration is still missing and the front desk is still empty.

Li Xiaomeng is still the same, still serving as the sacred and great patron saint of the tea room, playing games, watching TV series and reading comics all day long and having a silly time.

However, there was a rare guest in the tea room on the third floor today.

It’s Xu Qiu Sheng.

As soon as he came in, Xu Qiu Sheng was holding a document and saw Li Xiaomeng "investigating the market and gathering intelligence on friendly companies", and he couldn't help laughing.

This guy who is slacking off and playing games doesn't avoid people at all.

However, everyone in Fanxing knows that Director Li is second only to the first person in the studio, and everyone regards Li Xiaomeng as the second-in-command of the studio.

Don't be fooled by his silly appearance, but he has developed "Brave Heart" and "RPGmaker". Whether in terms of seniority or achievements, he ranks at the top in the entire studio.

"Director Xiao Li, I put the file there. Please tell General Manager Gu when she comes back."

"okay!"

Li Xiaomeng made a V sign to show that she understood.

After half an hour, Li Xiaomeng was tired of playing and got up to get snacks and drinks.

When I came back, I saw the documents on the table again and I became curious.

This is not something confidential, it is just a normal monthly report that briefly lists the development status of "Baldur's Gate" and the upcoming EA version of "Baldur's Gate".

Similar to the attitude of Development Team 1 towards Battlefield, Development Team 2 did not plan to conduct public testing.

But the difference is that this test of Baldur's Gate is a long-term open paid test with no limit on the number of places.

In addition to simply pulling the game out to see if it works, Development Team 100 also has the intention of crowdfunding. Due to the size of Baldur's Gate, there is too much content to be developed. In addition to expanding Development Team to people, a large number of outsourced projects are also required.

Not to mention that you will have to buy advertising space on major websites for saturated promotion later.

Even if "Victoria" and "Battlefield" made Fanxing a lot of money, this amount of money is still a huge sum.

The Linlang Project has invested tens of millions of dollars, and after several rounds, it will be able to squeeze out all the profits of "Battlefield".

Therefore, Development Team 2 is also trying to increase revenue and reduce costs.

Gu Qingshi had mentioned the EA version test to Xu Qiusheng before. She came back at night, read the document, and casually changed the chapter to pass it.

All that was left was to ask NGI's Chief Editor Lin to get a recommendation column on the platform - but unexpectedly, this was actually a bit difficult to do.

Gu Qingshi called Chief Editor Lin and said unhappily:

"Editor-in-Chief Lin, you promised me before that you would definitely get the recommended position!"

"Oh my goodness, I didn't expect Panshi Network to choose this time to release a new work, and even find our NGI to cooperate..."

Editor-in-Chief Lin was also at a loss whether to laugh or cry.

Speaking of this, we have to mention "Assault" and Panshi Network, which made their mark at the World Game Expo.

As the only American game developer at the exhibition, its special features need no further explanation.

Citigroup has always been at odds with domestic companies and has always been singing a different tune, but Citigroup's companies are eyeing the domestic market and often pretend to obey. The same is true for Panshi Network.

If it weren't for the market and sales, would they have bothered to attend the exhibition?

The authorities also attach great importance to bilateral trade and have been looking for opportunities to break the ice. The participation of Panshi in the exhibition and its willingness to cooperate are very good signals.

So when Panshi brought its new work "Lexington" to the NGI platform and wanted to release it in China, the top management of NGI had to give Panshi some preferential treatment and convenience based on political considerations.

In this case, it would be bad to let the stars face off against the mighty rock.

Therefore, Chief Editor Lin wanted Gu Qingshi to think about it again, and it would be best to separate it from "Lexington" to avoid a collision.

But Gu Qingshi puffed out her chest slightly, showing that she was not afraid at all, "No, I have every confidence in our game!"

"Little ancestor, have you really thought it through? Panshi is a big shot, with overwhelming publicity in December. It's not shameful to postpone the release..."

Editor-in-Chief Lin hurried to persuade him.

"I won't!" Gu Qingshi said in a very confident tone: "Panshi will release a new work in December, so we will conduct EA testing in November!"

Editor-in-Chief Lin: “…”

Good man, you sound so powerful, I thought you were really going to fight hard...

Just like that, the launch of the EA test of Baldur's Gate was finalized.

Gu Qingshi stretched and was planning to lie on the table for a while and then run to the tea room to use Li Xiaomeng as a human hot water bottle to warm her palms.

But unexpectedly, when things on this side were over, things on the other side started again.

But it was An Yan who called.

Gu Qingshi was also a little confused. It was already past nine o'clock in the evening, and sister, are you still thinking about work?

"Sister An, why are you calling so late?"

"Something went wrong on the battlefield. I was unable to contact you, boss, during the day."

Gu Qingshi became nervous for a moment and asked, "Is the problem serious? Is there some malicious bug?"

Malicious bugs and loopholes are not to be underestimated. Especially in an online game like Battlefield, if there is a mistake in the Shit Mountain program when updating the version, such as infinite props brushing or extra recharge rewards not being credited, then the game is over!

"That's not it. The number of cheating players has obviously increased during this period."

Gu Qingshi understood and immediately asked, "Is it because of the influx of new players during the Mid-Autumn Festival and National Day?"

As the two holidays coincided, the platform issued 88-38 discount coupons to players in order to promote consumption, and the sales of "Battlefield" also increased accordingly.

But the disadvantages are also obvious. The new player group is more mixed. In October alone, the incidence of cheating players has increased significantly.

There are also players in the game community complaining about occasionally encountering cheats, and these are small cheats with micro-auto-aim and little transparency. If you don’t watch the screen recording several times, you can’t recognize it.

The main reason is that the cost of cheating is too low.

NGI account registration is free, and the Battlefield game itself costs 138 yuan, which means the full discount coupon only costs 100 yuan.

One hundred bucks may not be as expensive as a plug-in!

Who cares about a hundred yuan when you can afford cheats these days? The elder brother was so happy that he gave several hundred yuan in tips to the female Bodhisattva in the live broadcast room.

"Let's do a cleanup first. I remember there are several companies that specialize in anti-cheating programs. Let's find one with a good reputation and use it first. As for how to solve the problem specifically... well, I'll think about it."

Gu Qingshi also had a headache about the problem of plug-ins.

The anti-cheating alliance consisting of software, hardware and players is still nowhere to be seen.

Hardware anti-cheating, Fanxing has just got the entry ticket to the peripheral market.

As for software anti-cheating, Fanxing’s current self-developed capabilities are not enough.

In terms of software anti-cheating, the most cost-effective anti-cheating method nowadays is disk scanning, which is convenient, saves money, and is cost-effective.

But the most accurate and primitive method is actually manual screening.

——As long as there are enough people and energy, we will definitely be able to identify the cheats.

The mouse movement and keyboard input of a cheater are very different from those of a normal player. If you can't see it once, you can always find the clues after watching it ten or a hundred times.

But this would make the labor costs too high.

So is there any anti-cheating method that is cheaper than manual screening and can have the same effectiveness?

Gu Qingshi thought about it and realized that there really was one.

Chapter 139 [AI Anti-Cheat System]

What is the AI ​​anti-cheating system?

This is an anti-cheating theoretical model that emerged in Gu Qingshi's previous life with the emergence of GPT. It is also the envisioned application method of deep learning models in the anti-cheating field in the future.

Simply put, it is to extract a large number of behavioral features of cheating players for AI to learn, and eventually make AI reach the level of being able to distinguish players' cheating behavior.

Although it is easy to say, it is actually very difficult.

Taking FPS games as an example, where do the large amounts of cheating data and normal data come from? If it is produced by an anti-cheating company, will the game company that provides the data sell the relevant data? If so, how much data is needed, and does this involve infringement of players' rights?

If a game company were to develop an AI anti-cheating system, would its self-development capabilities and financial reserves support it? Would the data from a single or just a few games be enough to fully train the AI?

After AI is trained, how should it be used? At the same time, in FPS games, will there be misjudgments due to differences in gameplay? If misjudgments or even a large number of misjudgments occur, who will be held responsible for these errors?

There are countless problems, unless a highly intelligent and strong AI emerges that can, through deep learning, have algorithms that are extremely close to human logic.

This kind of AI also exists in this world. Its prototype even appeared thirty years ago. It was initially used by the Soviet Union for its own OGAS (National Accounting and Information Processing Automated System), and it is also the basis of today's AI.

Thanks to the OGAS program, AI has made rapid progress, but it has never reached the level of strong intelligence.

It wasn't until 20 years after the collapse of the Soviet Union and two years ago that this thing was finally developed.

But unfortunately, this thing is a patent of the United States...

As for the domestic situation, after absorbing part of the Soviet legacy, rapid progress has been made in this regard.

The Oriental United University where Gu Qingshi worked was a higher education institution jointly sponsored by the domestic authorities and the Soviet Union at the time. The School of Artificial Intelligence of the Faculty of Information Engineering was doing this.

As an outsider, she didn't know the specific progress.

Let’s put the AI ​​anti-cheating system aside for now.

Ten days later, the news of the EA test of "Baldur's Gate" was released, and the pre-order page of "Baldur's Gate" could already be seen on the NGI platform.

In three days, the first public test of Baldur's Gate will be launched to players.

It has been six months, nearly half a year, since the release of Battlefield.

The dormant enthusiasm of many Star players was rekindled at this moment.

Many players have started to popularize the game's worldview in the NGI community - now DND is really not unknown at all.

The University of Michigan Library’s out-of-circle;

The DND board game appears;

Gu Qingshi spent money to put several DND-themed online novels on the bestseller list;