Phil Spencer, head of Microsoft’s gaming division, said in a company video that the decision is “not a change to our kind of fundamental exclusive strategy.”

Nevertheless, the shift signals a willingness to generate more revenue from content that previously could only be played on Microsoft’s own gaming hardware.

Microsoft completed the acquisition of prominent video game publisher Activision Blizzard for over $75 billion in October. In the fiscal second quarter, 11% of revenue flowing into Microsoft, the world’s most valuable public company, was tied to gaming. The Game Pass service that provides access to multiple games now has 34 million subscribers, up from 25 million two years ago, the company said. But the Xbox Series X and Series S consoles, which became available in 2020, have not proven as popular as the Sony PlayStation 5 or the Nintendo Switch.

“We think this is an interesting point in time for us to use what some of the other platforms have right now to help grow our franchises, so we’re going to do that,” Spencer said in the video.

Microsoft declined to identify the titles in question, but Spencer said they won’t be Starfield, a long-awaited game that came out in September, or the forthcoming Indiana Jones and the Great Circle.

Spencer did say two of the four are “community-driven games,” while the other two are smaller titles that weren’t meant to be exclusive to Microsoft’s own systems.

The Verge reported last week that Microsoft was looking into releasing the Indiana Jones game, previously announced for Xbox and Windows, on Sony’s PlayStation 5 in addition to Microsoft’s own console and PC operating system. The title will come from Bethesda Softworks, a subsidiary of ZeniMax Media, which Microsoft acquired for $8 billion in 2021.

Many Xbox fans expressed frustration on social media, speculating that Microsoft will give up trying to keep exclusive games on its console in the future. Without special content that can’t be played elsewhere, some argued, there would be less of a reason to keep investing in Xbox hardware and software.

One day after the Verge article was published, Spencer wrote in a post on X, “We’re listening and we hear you. We’ve been planning a business update event for next week, where we look forward to sharing more details with you about our vision for the future of Xbox. Stay tuned.”

Earlier this week, The Verge reported that games such as Hi-Fi Rush, Pentiment and Sea of Thieves would be coming to non-Microsoft consoles.

“I do have a fundamental belief that over the next five or 10 years, exclusive games, the games that are exclusive to one piece of hardware, are going to be a smaller and smaller part of the game industry,” Spencer said during the video appearance.

"Stress response" refers to a series of reactions produced by an organism in order to maintain homeostasis when it encounters pressure or threat from the external environment. It is a natural response produced by organisms to adapt to their environment and ensure survival. This response can be transient or prolonged.

On July 26, OpenAI's official Twitter announced that ChatGPT for Android is now available for download in the U.S., India, Bangladesh, and Brazil, with plans to roll out to more countries in the near future.ChatGPT is expanding its channels to gain more users and stronger stickiness in its use, as the wave of generative AI continues to push higher.

At the 2023 World Artificial Intelligence Conference (WAIC) in Shanghai in early July, a technician from a big-model startup was making her way through the exhibition hall, planning to find a cost-effective domestic chip solution for her company to use for big-model training.

"We have 1,000 A100s, but they are totally insufficient." She told HuffPost.

A100 is a high-end GPU from NVIDIA and the hardware base for ChatGPT growth. Some public data shows that the number of NVIDIA GPUs used by OpenAI in the process of training the GPT series of models is about 25,000 units. As a result, it has become almost customary in this industry to evaluate how many A100 graphics cards you can get in order to make a big model.

Where are the GPUs and where is the cheap arithmetic? This is just a snapshot of the many questions that will be asked at the 2023 WAIC conference.

All the people who have been "stressed" in the past six months are almost eager to find more answers about AI in this "event".

The technical staff of a chip exhibitor told Tiger Sense that in the past few days of WAIC, there were many product managers in front of their "big model" booth, who hoped to find product definitions for the company's big model business here.

At the Zhongguancun Forum on May 28, the China Institute of Scientific and Technological Information released the "China Artificial Intelligence Big Model Map Research Report", which showed that as of the end of May, 79 big models with parameter scales of more than 1 billion were released in China. In the following two months, there are Ali Cloud's Tongyi Wanxiang, Huawei Cloud's Pangu 3.0, and a series of AI models such as "ZiYao", etc., and according to incomplete statistics, there are currently more than 100 AI models in China.

Domestic enterprises scrambled to release AI model action, is the "stress reaction" best embodiment. The anxiety brought by this "reaction" is being transmitted to almost all the relevant personnel in the industry, from the CEOs of Internet giants to the researchers of AI research institutions, from venture capital fund partners to the founders of AI companies, and even many AI-related legal practitioners, as well as the regulatory layer of data and network security.

For those outside the industry, it may be a short-lived revelry, but how many people dare to say they are outside AI these days.

AI is ushering in a new era, and everything deserves to be reinvented all over again with big models. More and more people are starting to think about the consequences of the proliferation of technology.

Funds are pouring in and the flywheel is in place

ChatGPT was born in a month, the door to ask the founder Li Zhifei two to Silicon Valley, everyone must talk about the big model, in a conversation with Huffington Post, Li Zhifei bluntly said that this is his last "All in".

In 2012, Li Zhifei founded out the door to ask, this voice interaction, hardware and software as the core of the artificial intelligence company experienced the ups and downs of China's two waves of artificial intelligence. In the last wave of artificial intelligence in the hottest period of time, the valuation of the door to ask once pushed to the level of unicorn, but since then also experienced a period of despondency, until the emergence of ChatGPT, only to the silence of the artificial intelligence industry for many years to tear open a mouth.

In the primary market, "hot money is pouring in."

This is the industry consensus when talking about big models in the past six months. Qiqi Lu, founder of Qiqi Forum, believes that the AI big model is a "flywheel", and the future will be an era of ubiquitous models, "this flywheel has been started", and the biggest driving force is capital.

In early July, the business information platform Crunchbase released data showing that companies categorized as AI raised $25 billion in the first half of 2023, accounting for 18% of global financing. While this figure is down from the $29 billion raised in the first half of 2022, the total global funding across all sectors in the first half of 2023 was down 51% year-on-year from 2022, which shows that the AI sector's share of total global funding has almost doubled.Crunchbase wrote this in its report, "if there was no ChatGPT-induced AI boom, funding in 2023 would have been even lower."

By far, the largest funding round in the AI industry in 2023 was Microsoft's $10 billion investment in OpenAI in January.

Tiger Wink based on public data statistics, in the United States, the big model company startups, Inflection AI may become the second largest startup in the field of artificial intelligence financing after Open AI, after which are, respectively, Anthropic (1.5 billion U.S. dollars), Cohere (445 million U.S. dollars), Adept (415 million U.S. dollars), Runway ( 195.5 million dollars), Character.ai (150 million dollars) and Stability AI (about 100 million dollars).

In China, there were 456 public investment and financing events in the domestic AI industry in the first half of 2023. This compares to 731, 526, 353, 631, and 648 for the five-year period 2018-2022.

Another event that triggered the flywheel was the release of ChatGPT's API interface. When OpenAI first opened ChatGPT's API interface in March, there was almost a consensus within and outside the AI industry that the industry was about to change. With more applications accessing the big models, a more luxuriant forest is growing on top of AI.

"Doing the big model and doing the application itself should be separated", the investors' sense of smell is always sharp, in the view of Runze Chen, Executive Director of Source Capital, AI is a logic that is the same as the semiconductor division of labor, and the boom of the big model of AI will be followed by a wave of prosperity of AI applications soon.

At the beginning of this year, when Chen Runze traveled to Silicon Valley with his colleagues, he found that half of the projects of Y Combinator, the prestigious startup incubator in Silicon Valley (OpenAI CEO Sam Altman had served as president of this incubator for many years), had transformed to do generative AI. The enthusiasm for big models is no less than in China across the ocean today.

However, he also found that in the United States, whether it is capital or entrepreneurs, compared to the big model startup, more optimistic about the ecological applications based on the big model, after all, in this track, has run out of companies such as OpenAI, at the same time, the United States has a very strong ecological soil of the ToB application, so more U.S. companies are trying to do enterprise applications based on the ecology of the big model.

Chen Ran, co-founder of OpenCSG, a big model service platform, told TigerView that more than 90% of companies in the Bay Area have already used big model capabilities in every aspect. As for China, Chen Ran believes that before the end of the year a lot of customers will also be used.

Around March this year, Chen Ranze and his team began to try to find companies in China that do applications based on big models, but he found that there are very few such companies. A large amount of capital has entered the artificial intelligence industry, but if you trace the flow of these funds you will find that more money is still concentrated in the first few companies.

"Even now, it's already hard to invest in 1-2 out of 10 projects related to generative AI." In addition to Source Code Capital, Tiger Sense also communicated with a number of hard-tech investors, who all said that although the project to see a lot, but the truly reliable rare.

This attitude of the application side, in the view of many industry insiders, has been the norm.

Spectrum co-founder Yu Kai believes that the surface seems to be a lively track, in fact, more nominal competition, the result is no more than two situations: "One is to melt money, pure capital-oriented; the other is to do the whole domain of the general large model of the company, really need to shout, do not shout, if others will not know."

Some domestic statistics are also illustrating this problem, according to the third-party organization Enniu data statistics, as of July 2023, there are 242 AIGC companies in China, and there have been 71 financing events in the AIGC track since January. In contrast, there are 67 companies on the AI big model track, with only 21 financing events since the release of ChatGPT.

The earnings report shows that Meta's total revenue in Q2 2023 was $32 billion, up 11% year-on-year, exceeding the market's expectation of $31.06 billion; Q2 earnings per share were $2.98, up 21% year-on-year, the highest since Q4 2021, compared with the expectation of $2.91; net income was $7.788 billion, compared with the market's expectation of $7.504 billion and $6.687 billion in the year-ago quarter, up 16% year-over-year.

This is the first time Meta has seen double-digit revenue growth since the fourth quarter of 2021.

After a full suspension of the Meta universe, Zuckerberg finally admitted defeat and centered the business on the social space and AI, giving Meta a nirvana.

Three consecutive years of losses for Meta Universe

The Metaverse program has brought Meta tens of billions of dollars in losses, and that number is still growing.

The Reality Labs division, where Meta Universe is based, lost $3.73 billion in the Q2 2023 quarter, up from an expected $3.68 billion and up from $2.8 billion a year earlier, and $3.992 billion in the Q1 quarter, according to the earnings report.

Even Meta's Metaverse division has lost more than $10 billion for two years in a row, totaling $13.717 billion in 2022 and $10.193 billion in 2021.

In other words, since Zuckerberg's "All in Meta Universe", Facebook has already lost about $31.6 billion in the Meta Universe program since it changed its name to Meta, and its stock price has lost about two-thirds of its value directly.

In his 2022 earnings call, Zuckerberg reluctantly admitted his failures and promised that 2023 would be "the year of efficiency," with tight cost controls, cuts to middle management and underperforming programs. And he did keep his promise, taking the lead on the Metaverse program.

Meta first called a halt to production of the VR device, no new parts will be ordered for the Quest Pro, and Gore, which is in charge of assembly, will only produce the Quest Pro as long as the current number of parts allows.

And Meta began to transform the meta-universe game, set up Ouro Interactive's internal studio, established the "Horizon Worlds" (Horizon Worlds) project, launched VR virtual games, trying to use the meta-universe to make some "fast money The project was an attempt to make some "quick money" from the metaverse.

After stopping the loss in time, Zuckerberg looked to the hottest project today, generative AI.

At the beginning of March this year, Meta officially released Llama, a large-scale language model, which has been widely praised in the AI community for its small size and strong performance, and is considered an important tool for advancing AI research and development.

But this is not the "big move" that Zuckerberg has been waiting for.

In the early hours of July 19 this year, Meta released a large language model Llama 2, compared to Llama 1, the new version of Llama 2 training data volume increased by 40%, and open the corresponding chatbot bars have version Llama 2-Chat.

Moreover, Meta announced that the Llama 2 model is an open source and free commercially available large language model, breaking the existing specifications of AI in one fell swoop and becoming the first free, open source, commercially available generative AI large model.

On Facebook, Zuckerberg called the Llama 2 model the next generation of big models.

Unlike closed-source models such as ChatGPT, GPT-4, etc., Llama 2 openness attracts a very large number of users, and anyone can request to download Llama 2 by filling out a form on Meta's website.According to the official news, since its release, there have been more than 150,000 requests for Llama 2 downloads in a week.

A hit, Llama 2 became the hottest open source ChatGPT alternative. OpenAI, which has been in a dominant position, has been beaten back by the "free" flag.

Socialization is still the basic plate

While Meta has made a name for itself in the field of generative AI, its fundamental disk remains the social space.

In early June, Meta held a staff meeting to show a social app called "Threads", the page layout and operation is very similar to Twitter, Threads is planned to be released in early July, and invited a number of celebrities and netizens to test the application, including "" talk show queen" Oprah Winfrey (" Oprah Winfrey ")," and "" the social media" Oprah Winfrey (" Oprah Winfrey "). talk show queen" Oprah Winfrey.

Threads instantly infuriated Twitter boss Musk, who directly rebuked Zuckerberg on Twitter, and threatened to "fight" Zuckerberg in the Octagon. Zuckerberg directly personally down, agreed to Musk's fight.

Whether it's the tech world bigwigs, or the media world on this $340 billion showdown big hype. Although this "fight" finally ended, but Zuckerberg successfully utilized the traffic brought by this marketing for Threads, attracting a large number of users to pay attention to the launch of Threads.

After the launch of Threads, it immediately set a number of records for social apps: 5 million registered users in 4 hours; 10 million in 7 hours; 30 million in 24 hours; and 70 million in 48 hours....... On the fifth day of the official launch of Threads, the registered users of Threads exceeded 100 million.

In contrast, it took five years and two months for Twitter to surpass 100 million registered users.

In addition to the number of users continues to wildly, more rare is Threads users are very active, according to The Verge statistics, the number of posts on Threads in the past few days has reached more than 95 million, the number of likes 190 million times.

At the same time, many brands have also begun to "settle" Threads, including Netflix, Nike, Adidas, L'Oreal Paris and other top global brands. Even Disney, Twitter's biggest advertiser, has opened an account on Threads.

The fact that advertisers are all moving into Threads means that Twitter's biggest source of revenue, ad dollars, have been slashed. Musk was so furious that he sent a cease-and-desist letter to Zuckerberg, threatening to sue Meta for "systematic, willful, and unlawful misappropriation" of Twitter's trade secrets and intellectual property, as well as for stealing Twitter's data.

On the surface, the explosion of Threads was a coincidence of factors. But behind the scenes is Zuckerberg "planning" for many years, day and night thinking about the Twitter market ambition.

As early as 2008, a tentative offer of 500 million to start, has been trying to buy Twitter. Later acquisition failed, Zuckerberg on the use of their own "good to learn from" means, the successive launch of text-based social platforms, to seize the Twitter market.

First, Instagram shelves "Notes" new function, so that users can publish 60 characters of short tweets through this function; and then plans to launch "Project 92" App to compete with Twitter, the App will be The app will be accessible through an Instagram account.

It can be said that before this, Zuckerberg has tried countless times to imitate Twitter, and finally with Threads to fulfill the long-cherished dream, to realize the occupation of the Twitter market.

The meta-universe is still sucking blood

Meta has won back users with its Llama 2 and Threads apps and is gradually increasing its profitability through major layoffs.

As of Q2 2023, Meta's total workforce was down 14% year-over-year to 71,469, according to the earnings report. The composition of the workforce has shifted toward higher-cost tech roles and plans to invest more in payroll expenses.

But the meta-universe division is still still sucking blood out of Meta, and despite Meta's outward claims that it has slowed down meta-universe research and development, Meta still paid $3.73 billion and $3.992 billion in bills for meta-universes in 2023.

Horizon Worlds, the centerpiece of Meta's meta-universe business, continues to fall in subscriber numbers, and at the end of 2022 even had fewer than 200,000 subscribers, putting it in a decidedly backward position.

Spending the most money and taking the most vicious beating is a perfect way to describe the Meta Meta Universe business.

After half a year of honest development, Meta has forgotten the pain, not only this year's "year of efficiency" expectations have not yet been realized, and began to make aggressive spending.

According to MetaQ2 quarterly earnings call, Zuckerberg plans to further invest heavily in virtual worlds and artificial intelligence in 2024, and listed three plans to significantly increase spending in 2024. These include increased spending on depreciation expense, spending on personnel in highly technical positions, and spending on the Reality Labs division.

Accumulated losses of $ 31.6 billion meta-universe project, still did not stop Zuckerberg increased support for Reality Labs division. This makes many foreign analysts bearish Meta's subsequent development.

After two years of consecutive losses, Zuckerberg and Meta have learned to cut costs and efficiency, and took advantage of the generative AI big model and the explosion of the social field of class tweeting, ushered in the first double-digit revenue growth.

Zuckerberg never let go of "All in Meta Universe", whether it will fall back into the Meta Universe trap, is still an uncontrollable factor.

On the other hand, if Meta can really realize the meta-universe, generative AI, social three areas together, then it is expected to regain its former glory, once again to catch up with Apple, Google, Microsoft's footsteps.

In addition to shareholder holdings, AI hot also triggered a wave of divorce of the actual controllers of listed companies, according to statistics, just over half of this year, AI has been "broken up" Fubon shares, three six zero, Tongcheng new material, science and technology, Tiandi digital, Tonghe technology, back to the new material, Saiteng shares, Jusheng micro 9 major shareholders of the family.

It is worth noting that the above divorce announcement and the male party "net out of the household", the female party to take the stock, which can not help but let the stockholders have a guess - divorce is fake, holding is real.

The shareholders have also spit: AI fire "side effect" is actually divorce?

AI wealth behind the tide of holdings, A shares divorce "curve to save the country"

Since the beginning of the year ChatGPT a hit, AI has set off a wave of investment, all kinds of funds have poured in, domestic related concept stock prices climbing. As of June 20 (the high point of the year), Longchao information has risen 202.04%, the share price has doubled. Kunlun Wanwei shares also rose from 14.62 yuan to 62.93 yuan, an increase of 336.17%, the highest share price in the year had come to 70.66 yuan. In addition, Shengtian network, Shenzhou Taiyue rose by more than 2 times during the year.

Looking at the world, the market value of a number of artificial intelligence companies is rising. As of June 23, the closing of U.S. stocks, AI big brother Nvidia shares have risen 188.90% so far this year, the market value has exceeded $1 trillion. In addition, BigBear.ai has risen 247.18% so far this year, Oracle, Google rose 46.29%, 38.66% respectively.

In the face of the market value soaring, a large number of AI stocks large shareholders began to choose to reduce their holdings.

Just last week, Harvey Jones, a long-time board member of Nvidia, sold nearly 120,000 shares of Nvidia stock, cashing out $48.3 million. He has already cashed in over $76 million this month. By coincidence, TenchCoxe, a member of Nvidia's board of directors, also sold 50,000 shares of Nvidia stock at an average price of $422.15, making a profit of more than $21 million.

Oracle founder Larry Ellison also chose to cut his holdings at the high point to cash out, recently exercising his expiring options to buy 5.25 million shares at $30.11 per share. At the same time, Ellison sold the same number of Oracle shares at an average price of $121.98 per share, for a total of $640 million.

Compared with the simple and brutal reduction of overseas AI listed companies, A-shares are a "curve to save the country".

This reporter found that since this year, A shares have 9 AI concept stocks have disclosed the news of major shareholder divorce, including Fubon, three six zero, Tongcheng new material, science and technology, Tiandi digital, Tonghe technology, back to the new material, Saiteng shares, Jusheng micro, some of these companies have attracted public attention because of the "divorce case of the sky" and in the The public opinion field set off an uproar. It is worth noting that the timing of the divorce of the above shareholders is basically when the company's share price is soaring.

For example, Zhou Hongyi, the actual controller of Three Six Zero, split 44658.52 million shares (about 6.25% of the company's total share capital) to his ex-wife Hu Huan's name, the closing price of 20.08 yuan at the time of the divorce (April 4), which is the highest point of Three Six Zero in the year.

In the divorce case of Tongcheng New Material, the man was directly "netted out" and the shares with a market value of over 14 billion yuan went to the woman.

Recently, Jusheng Micro's controller Tang Zhuang transferred 32,757,500 shares (about 6.14% of the company's total share capital) to his ex-wife Yi Gobing's name, and the closing share price was 104.12 yuan on that day, which is also experiencing a rise after the adjustment of artificial intelligence.

The stockholders have been complaining: AI wave brought about by the "side effects" is actually divorce?

In addition, the Dragon Boat Festival before the collapse of the AI sector and Kunlun Wanwei, its announcement that its second largest shareholder, founder Zhou Yahui's ex-wife Li Qiong plans to reduce its holdings of no more than 3% of the shares, but also to support the long-term development of the company's AIGC business, but also to reduce the shares of more than 50% of the after-tax proceeds of the loan to the company.

It is understood that Li Qiong and Kunlun Wanwei founder Zhou Yahui have divorced from each other in 2016, when Li Qiong split to obtain 278 million shares of Kunlun Wanwei.

There are also market participants jokingly said, "it is recommended to watch the software to increase the 'divorce concept shares'."

Wu Gao Bin, vice president and secretary-general of the founding committee of the CNA meta-universe worker, told reporters that "divorce is a private matter in the life of an individual, but as the AI company's major shareholders divorce on the company's share price and business development may have an impact. Some divorces may lead to instability in the company's shareholding and business development direction, which may cause market concerns and investor uneasiness, which in turn may lead to share price volatility."

How to respond well to market capitalization growth?

For the market "fake divorce really reduce" speculation, a market participant expressed doubts, "the matter can not be simply blamed on the market, if in fact the feelings have long broken, why not in the previous decision to make a divorce? It is really puzzling to choose to pay billions of dollars of separation fees at a time when the stock price is soaring."

Zhang Zhuran, an independent economic researcher, said in an interview with Caixin, "If the divorce was indeed chosen due to personal emotional problems, there is nothing wrong with that. After all, emotional problems and marital choices are personal matters, and there is no need for outsiders to say anything. However, the coincidental timing of the divorce did cause concern among investors, after all, there may be considerations of holdings and profits."

All the above analysis shows that both A-shares and overseas, the share price surge is commonly accompanied by a reduction in holdings. For this, how should listed companies do a good job of market value growth response?

Market participants generally believe that market value growth needs to rely on the fundamentals, rather than the illusory wind and speculative concepts. If the industry bubble bursts, only a few enterprises that master the core technology can survive.

Zhang Zhuran pointed out that "the basis for listed companies to do a good job of market value management is first of all to do a good job of their own management and strengthen the 'fundamentals' of listed companies, which is the primary work, the growth of market value is not only a true response to its fundamentals, but also a concrete reflection of the heat of the market. "

Wu Gaobin holds the same opinion, he believes that market capitalization growth needs to be supported by the inherent strength of the company, including a high-quality business model, strong technological innovation capabilities, an efficient management team and sound investment strategies. Excessive focus on increasing market capitalization may result in companies focusing too much on short-term performance and neglecting investment in long-term development, thus reducing future competitiveness and market capitalization. Therefore, the Company should adopt a prudent business strategy and consider various factors in order to avoid financial and operational risks.

A market researcher also confessed to the reporter, "Although the actual controller is a fake divorce lack of conclusive evidence, but in the A-share market, 'fake divorce, real holdings' has become a stress judgment, shareholders are forced to become a major shareholder divorce 'with the gift party '."

From the secondary market performance can also be a glimpse of two, in the divorce or holding announcement issued, the "divorce concept stocks" share prices are not able to withstand the test. Such as three six zero share price increase stopped at the highest point on April 4, Kunlun Wanwei ex-wife reduction announcement also brought a collapse of a number of AI concept stocks. This side alert investors, the current domestic AI concept stock bubble is large, the announcement of holding reduction intensive release, also indicates that the shareholders themselves also think that the stock price is overvalued, appropriate stay away is the right way.

For the response after the growth of market value, Zhang Zhuran rhetorically asked, "If you respond with 'divorce' and 'holding reduction' as soon as the market value grows, it only proves that the relevant stakeholders themselves are not optimistic about the prospects of the company, and how can this contribute to the long-term confidence of the market?"

Wu Gaobin also said that in dealing with market value growth, it is also necessary to comply with relevant regulations and market rules, regularly disclose financial reports and business plans, and establish a strict monitoring mechanism and risk management system in terms of internal control and auditing to ensure sustainable development of the company and a solid increase in market value. In addition, in the event that non-operating factors such as divorce of major shareholders have an impact on market capitalization, companies need to actively take countermeasures to reduce risks and stabilize market sentiment and investor confidence by strengthening dissent management, improving corporate governance and other measures.

The above-mentioned market researcher frankly said that in terms of the capital market, as a place for resource allocation, it should also respond to market capitalization growth and hot phenomenon in a sensible manner. At present, some artificial intelligence concept stocks in the A-share market have P/E ratios as high as 100 times or 1,000 times, even if the domestic AI sector is developing rapidly again, it is less likely to fill such a huge valuation bubble, which may lead to a mismatch of resources.

There is also the view that the increase or decrease in holdings is the shareholder's independent behavior, under heavy restrictions, often divorce and other deformation of the move.

The first quarter of the public funds into the field significantly, the second quarter will also reduce positions?

The rise of artificial intelligence is also attracting the inflow of institutions and funds, and public funds are no exception.

Wind data shows that, ranked by the total market value of holdings, the top ten long positions of public funds in the first quarter of this year were Guizhou Maotai, Ningde Times, Wuliangye, Luzhou Laojiao, Tencent Holdings, Wu Mingkangde, Myriad Medical, Shanxi Fenjiu, Jinshan Office and Sunshine Power.

Among them, Jinshan Office for the first time in the past three years into the top ten long positions of public funds, holding a total of 67,088,500 shares, becoming the "epitome" of public funds layout AI. Such as Manulife Transformation Opportunity A, Boshi Hui Xing return a year hold, Efatar CSI Venture 50 ETF, Fortune New Power A and other funds are new to buy Jinshan Office in the first quarter.

In addition, in the first quarter of the top 100 long positions in public offerings, ranking up more than almost a clean sweep of technology stocks, such as Kunlun Wanwei Haikang Weiwei, in micro companies, Hang Seng Electronics, KDDI, ZTE, Cambuji, etc..

Kunlun Wanwei, for example, Kunlun Wanwei at the end of 2022 only 12 index-based funds holdings, the number of shares held 18.21 million shares. By the end of the first quarter of this year, the number of public funds holding Kunlun Wanwei surged to 136, and the number of shares held reached 64,226,800, an increase of up to 253% from the end of last year.

Choice data shows that AI concept companies are mainly distributed in computer, media (including games), communications, electronics and other TMT industries. As of the end of the first quarter, these four industries of A-share listed companies have public funds holdings of 626 companies, the total market value of fund holdings of these four industries is close to 600 billion yuan, including electronics, computer industry, held by the fund market value of more than 200 billion yuan.

Near the end of the second quarter, the fund manager's operation of the new trend has become a hot spot of market concern, is to continue to hold AI or escape the top? According to a sell-side sources, some fund managers also began to reduce their positions AI, choose to fall into the bag for peace. "The early AI rose too fast, with the rise in market value, fund managers may be worried about the wind after the relevant sector shock, so will choose to reduce positions in due course to ensure profit."

Some fund managers have been cautious about the AI sector, the pre-performance TMT sector still needs to go through a period of shock and digestion process to further reduce trading congestion and excessive market speculation, the process may involve a pullback in share prices, earnings adjustments and the calming of market sentiment. "We will allocate a moderate amount, but will not take a heavy position."

There is also speculation in the market that some style-drifting funds may take the opportunity to sell AI.

On Tuesday, Business Insider reported that Google terminated its augmented reality (AR) headset project, codenamed Iris, in favor of building an AR software platform.

Currently, the headset giants are starting to "go their separate ways", with Microsoft spitting out the VR and AR market as too small at the moment, and the XR business unit having been completely disbanded earlier in the year.

However, Apple and Meta are still insisting on MR. Earlier this month, Apple released its first headset "Vision Pro", officially entering the MR track; Meta, which was once all in the meta-universe, continues to invest heavily, but has switched its focus from the meta-universe to AI.

Google terminates AR headset project

Google terminated the project of manufacturing augmented reality (AR) headset and turned to the construction of AR software platform.

Business Insider reported that three people familiar with the matter revealed that the glasses, internally codenamed Iris, were put on hold earlier this year due to layoffs, reorganization and the departure of Google AR/VR CEO Clay Bavor.

The Verge first reported the existence of the Iris project in January 2022, saying the device looked similar to a pair of ski goggles. A Google employee pointed out that the ski goggles were actually a separate AR project, while Iris was a series of devices that were more like glasses.

Since shelving the Iris glasses, Google has been focused on building an AR software platform, and employees are developing software using a prototype platform known internally as "Betty," with one employee describing Google's ambition as "Android for AR ".

Microsoft trolls market size too small

Microsoft recently trolled the XR market is currently too small, and all XR business units were disbanded at the beginning of the year.

Recently, Matt Booty, the head of Microsoft's game studio Xbox Game Studios, pointed out in an interview that the virtual reality (VR) and augmented reality (AR) market is currently too small to support Xbox's game platform to get involved.

In January this year, Microsoft disbanded all of its XR business units involving metaverse such as HoloLens, AltSpaceVR, MRTK, and MWR, followed by the disbanding of the industrial metaverse team in February.

Meta wants to empower the metaverse with AI

Once All in Meta UniverseMeta still continues to invest heavily in Meta Universe and plans to integrate AI technology into Meta Universe in the future.

Meta previously preempted Apple to release a new generation of virtual reality headset Quest 3, the headset selected GPU chip, processor capacity and Quest 2 compared to double.

Zuckerberg switched the focus from the meta-universe to AI at the all-staff meeting, where he clarified the company's AI-focused development strategy, pointing out the future integration of AI technology with products such as the meta-universe and Instagram.

Apple releases heavyweight MR headset

In early June, Apple released its first headset, Vision Pro, at the WWDC conference, marking its official entry into the MR market.

Vision Pro supports gesture recognition, eye-tracking and other interactive methods, and is priced from $3499. Apple also released VisionOS, a new operating system built for space computing, at WWDC, which can support running various applications on iOS and iPadOS.

It is worth mentioning that, compared to the virtual world created by Meta VR, Vision Pro pursues the integration of reality and virtual, so that users will not be completely isolated from the surrounding environment when using it, greatly broadening the use of scenarios.

Apple CEO Tim Cook positioned this product as a revolutionary spatial computer device that allows users to operate apps, games, and movies on digital devices in real space, seamlessly integrating digital content with the physical world.

More than 600 of Microsoft's largest customers, including Bank of America, Walmart, Ford and Accenture, are expected to try out the artificial intelligence capabilities in Microsoft Office 365, and at least 100 of them have already paid an additional annual fee of up to $100,000 each for 1,000 subscription accounts, according to technology media outlet The Information. ($100 more per account on top of the original Office 365 annual fee)

Compared to the classic version, this AI version of Office 365 is at least 40% more expensive, with value-added features such as automatic text writing in Word documents and automatic PPT creation.

In May, Microsoft expanded its beta testing, inviting the first 600 enterprise customers worldwide to experience the feature, and added semantic indexing, integration of OpenAI's text-to-picture model DALL-E into PPT, quick drafting of email content and several other smart features.

One of the reasons for the higher pricing is that these features typically require servers with special chips that consume more power than traditional servers, according to a person with direct knowledge of the program who spoke to the media.

In addition, the person also said that the number of participants in the trial program has so far exceeded Microsoft's expectations.

Pricing exploration

The small-scale pilot will provide Microsoft with a pricing strategy to explore, as well as provide a reference for other software companies that incorporate AI features.

Microsoft has reportedly been weighing two potential pricing models for AI features when they become widely available: one is to charge for these features as an add-on to a basic Office subscription, and the other is to add AI features to Office for all enterprise users, raising the price of all subscriptions across the board when customers renew their subscription plans.

Last March, Microsoft raised the price of the basic version of Office 365 by 20 percent, with the most common subscription version now costing $22 per month. That means customers who currently buy 1,000 accounts on that version will pay about $264,000.

Of course, the bigger the customer the corresponding discount may be available.

High Costs

Cost is the main reason for Copilot's price increase: both the cost of training the model upfront and the cost of running the servers later.

Earlier in the day, Wall Street News reported that a significant portion of Microsoft's early investment in OpenAI was achieved with credits for the use of its cloud platform, Azure. And to go along with OpenAI, Microsoft also spent $1.2 billion to build a supercomputer to develop the models, and invested another $10 billion earlier this year - equivalent to about one-sixth of Microsoft's annual cash flow.

1. OKX

Official website link: https://www.okx.com/

OKX will launch SUI at 20:10 on May 3, with the opening of the trading session using pooled bidding and the opening of bidding at 19:50 on May 3, and will open SUI withdrawals at 20:15 on May 3. OKX Ventures, as a partner of Move Accelerator, has paid much attention to the Sui ecosystem and has participated in the investment of several Sui ecosystem projects.

1. Cryptocurrency

Official website link: https://www.binance.com/

Users can start pledging BNB and TUSD for mining on May 1, 2023 at 8:00 BST for two days. This time, CoinSec Launchpool will provide a total of 40 million SUI rewards, accounting for 0.4% of the total SUI supply. In addition, Cryptocurrency will launch SUI when liquidity requirements are met, and open SUI/BTC, SUI/USDT, SUI/TUSD, and SUI/BNB trading pairs.

1. Bitget official website link: https://www.bitget.com

Bitget announced that it will launch the Sui token SUI in the Innovation Zone and launch a free airdrop program for the whole network. There are two ways to participate in the airdrop, namely SUI official Discord certification and net top-up, with a total prize pool of 35,000 SUI.

1. Bitfinex

Official website link: https://www.bitfinex.com/

Bitfinex announced that it will soon launch Sui (SUI), and will soon announce its top-up and trading opening hours.

1. Coinlist

Official website link: https://coinlist.co/

CoinList announced in an announcement that it will launch Sui's native token, SUI, on the same day as SUI. At the same time, CoinList Pro will offer 48 hours of free trading for all SUI pairs and 1 month of free pledging to all eligible CoinList users.

1. Link to Kucoin's official website: https://www.kucoin.com/

Kucoin released an announcement that it will go live with SUI Token on May 3, and Kucoin launched the subscription mechanism of WL whitelist for Sui Token. As a partner of Move Accelerator, Kucoin pays much attention to Sui Eco and actively invests in Sui Eco projects.

1. Bybit

Official website link: https://www.bybit.com

Bybit released an announcement to announce that SUI Token will be online on May 3, and Bybit launched the subscription mechanism of WL whitelist of Sui Token, as the partner of Move Accelerator, Bybit pays high attention to Sui ecology and actively invests in Sui ecological projects.

There is no airdrop in Sui, but there is a token community access plan

There is no community airdrop plan clearly after Sui mainnet is launched, please users beware of fraud and phishing activities.

However, Sui has released SUI token community access plan to offer the right to buy SUI at low price to backers. Those eligible to purchase include the winners of the Capy Holidays contest during the 1st round of testing, some members involved in organizing Builder Houses and other important events, and some community participants. One of the most widely reached users is the Token Whitelist program for Sui Discord activists.

Eligible backers must be active on Sui Discord until February 1, 2023, and active users can submit information and have their eligibility verified in order to activate the whitelist. Whitelist users can subscribe to up to 1,500 SUI at a price of 0.03U starting April 23 at 10:00 and ending April 24 at 10:00.

The IEO for SUI is available on the three major exchanges and the details of the sale are as follows

KuCoin: 225 million SUIs (no support for mainland users)

Group A general public sale, KuCoin randomly takes a snapshot of the user's total USDT, USDC, ETH, KCS balance to determine the final number of tickets available to each user and selects the winner by lottery.

Group B identification sales, this round is only available to early contributors to Sui and must be a Sui whitelist user.

Bybit: 940 million SUIs (no support for mainland users)

Subscriptions on Bybit's token sales platform ByStarter

OKX: 225 million SUIs (supports mainland users)

Subscribe in JumpStart, the token sales platform of OKX or Sui whitelist users can realize low price purchase.

Those Sui ecology that you must pay attention to

Token distribution and airdrop plan

At present, there are some Sui ecological projects have announced financing or Token distribution plan.

1）Suia

Official link: https://suia.io/

Sui ecological social protocol Suia announces that it will issue Token SUIA, the total amount of SUIA is 100 million, 7% will be allocated to community access plan and unlocked immediately; 13% will be allocated to early contributors and unlocked linearly in 20 months; 20% will be allocated to team and unlocked linearly in 40 months; 60% will be allocated to community and unlocked linearly in 50 months. SUIA will be used for governance and user incentives.

1. Abyss World

Official link: https://www.abyssworld.games/

Abyss World, the first 3A open-world dark fantasy ARPG chain game built on Sui Network, announced that it will be dropping SUI and AWT Token to holders of "Gazer - Abyss World".

R-level NFTs can claim 400 SUIs and 1,600 AWTs; SR-level NFTs can claim 1,000 SUIs and 4,000 AWTs; SSR-level NFTs can claim 2,250 SUIs and 9,000 AWTs; SP-level NFTs can claim 12,500 SUIs and 50,000 AWTs. Token will be released for 10 months, and the specific release rules will be released at the end of May.

1. Cetus

Official link: https://www.cetus.zone/

Cetus announced on May 2 that it has completed a seed round of funding led by OKX Ventures and KuCoin Ventures, and led by Comma3 Ventures, NGC Ventures, Jump Crypto, Animoca Ventures, IDG Capital, Leland Ventures, AC Capital, Adaverse, Coin98 Ventures, and others.

4）Turbos Finance

Official link: https://turbos.finance/

Turbos Finance, a Sui eco-decentralized trading platform, announced that it has received investment from Jump Crypto for an undisclosed amount. Turbos Finance is also one of the first accelerated projects announced by Move Accelerator, more projects: Move Accelerator first accelerated projects announced, Let's Move!

Eco-projects on the main network

On the occasion of Sui mainnet's official launch, there are many ecological projects that have finished debugging and integration on the test network, they will be released together with the launch of Sui mainnet, the projects range from DeFi, NFT platform & game, Launchpad, wallet, infrastructure and other tracks.

The unique advantages and token economics of Sui

The total supply of SUI tokens is 10 billion, which will be allocated to 50% community reserve, 20% early contributors, 14% investors, 10% Mysten Labs vault, 6% community access program IEO and application testers. There are four main application scenarios for the tokens: network pledges, transaction fees, storage fees, governance voting, and SUI native asset trading tools.

Compared with other public chains, SUI's unique advantages are:

1. Through features such as horizontal scaling, composability, sparse replay, and on-chain storage, Sui's architecture solves a common pain point of Layer1.

2. On Sui Network, each set of transactions is processed in parallel, contrary to the bottlenecks that occurred in earlier blockchains due to the lack of distinction between various objects, resources, accounts, and other components.

3. Unlike most other blockchains, assets (e.g., NFTs) can be passed directly into function parameters in Sui Network. sui's object-centric approach also allows for more complex data structures and the ability to store assets within such data structures or within the assets themselves.

4. Blockchain provides a ledger of every transaction on the chain, but it is costly to query the data on the chain, so with SUI, game developers do not need to keep track of unrelated Dapp transaction records, and products will be able to follow the evolution of objects in this game without having to mine data from Merkle trees.

5. Since assets are stored directly on the Sui blockchain as objects, they are never bound by the Merkle tree index. Because it is much cheaper to update assets directly on the chain, storing assets directly on the chain in combination with traditional methods such as IPFS can solve the problem of high on-chain storage costs.

Next, we'll explore the ways in which female characters drive emotional connection in film, the importance of representation and diversity, and the steps you can take to get involved in our community and receive a free NFT by subscribing to our account.

The Impact of Female Characters in Film

Female characters have a unique ability to drive emotional connection in film. They can represent the complexities of human relationships and provide a lens through which we can view the world. Female characters can evoke empathy, compassion, and understanding, and can inspire audiences to connect with the film on a deeper level.

Female characters can also challenge traditional gender roles and stereotypes, and can provide representation and diversity in a male-dominated industry. When female characters are portrayed as strong, complex, and multidimensional, they can serve as role models for women and girls and inspire them to pursue their own dreams and ambitions.

The Importance of Representation and Diversity

Representation and diversity are crucial to the success of any film industry. When female characters are portrayed in a positive and respectful manner, they can help to challenge gender stereotypes and promote gender equality. Similarly, when female characters are diverse in terms of race, ethnicity, and sexual orientation, they can provide representation and inspire audiences from all backgrounds.

Steps to Get Involved

If you're interested in learning more about the impact of female characters in film and promoting representation and diversity in the industry, we invite you to join our community. Our account is dedicated to sharing our insights and analysis on the role of female characters in film, and we're always looking for new members to join our community.

For the suprise gifts, anyone who subscribed now can mint your own free NFT, which will give you exclusive access to our community and our insights into the impact of female characters in film.

https://opensea.io/assets/ethereum/0xBAd8072B617c2bbFB0715f926dd89Accc35e0fd0/0

In sum, female characters play a vital role in driving emotional connection and challenging traditional gender roles in the film industry. By promoting representation and diversity in the portrayal of female characters, we can inspire audiences and promote gender equality. Join our community to learn more and receive a free NFT as a special offer.

However, even on a strictly technical level, I have not yet succeeded in becoming a believer. So, in light of all the recent attention to what is now called web3, I decided to explore some of what is happening in the field more thoroughly to understand what I may have missed.

How I see web 1 and web 2

web3 is a somewhat vague term and it's hard to critically assess what web3's ambitions should be, but the general argument seems to be that web1 was decentralized, web2 centralized everything on the platform, and web3 will decentralize everything again. web3 should give us the richness of web2, but decentralized.

It might be good to have some understanding of why centralized platforms emerged, and in my opinion, the explanation is simple.

People don't want to run their own servers, and they never will. web1 is based on the premise that everyone on the Internet is a publisher and consumer of content, as well as a publisher and consumer of infrastructure.

We all have our own web servers and our own websites, our own mail servers for our own email, and our own state servers for our own persona generation. And yet - and I don't think this can be stressed enough - this is not what people want. People don't want to run their own servers.

Even nerds don't want to run their own servers at this point. Even organizations that are building software full time don't want to run their own servers at this point. If there's one thing I want us to understand about the world, it's that people don't want to run their own servers. The companies that provide you with these services are successful, and the companies that iterate on new features based on the possibilities of these networks are even more successful. Protocols run much slower than platforms. after 30+ years, email is still not encrypted; meanwhile, WhatsApp went from no encryption to full e2ee (end-to-end encryption) in a year. People are still trying to standardize reliable video sharing via IRC; meanwhile, Slack allows you to create custom reaction emojis based on your face.

It's not about funding. If something is truly decentralized, then it becomes very difficult to change and often gets stuck in time. It's a problem for technology because the rest of the ecosystem is evolving rapidly and if you don't keep up, you're going to fail. But web3 intends to be different, so let's take a look. To quickly understand the space and get a better idea of what the future might hold, I decided to build a couple of dApps and create an NFT.

Making some decentralized apps

To get a feel for the web3 world, I made a dApp called Autonomous Art that allows anyone to mint a token by making a visual contribution to the NFT. The cost of making a visual contribution increases over time, and the contributor's minted funds are distributed to all previous artists (visualizing this financial structure as similar to a pyramid shape). At the time of this writing, over $38,000 has been spent on creating this collective artwork.

I also created a dApp called First Derivative which allows you to create, discover and exchange NFT derivatives that track the underlying NFT, similar to financial derivatives that track the underlying assets.

Both give me some insight into how the space works. To be clear, there is nothing particularly "decentralized" about the applications themselves: they are just normal websites. "Decentralized" refers to where the state and the logic/permissions to update it are located: on the blockchain rather than in a "centralized" database.

One of the things I've always found strange about the cryptocurrency world is the lack of focus on the client/server interface. When people talk about blockchain, they talk about decentralized trust, leaderless consensus, and all the mechanisms by which it works, but often obscure the reality that clients ultimately cannot participate in these mechanisms. All network graphs are server-based, the trust model is server-to-server, and everything is about the server. The blockchain was designed as a peer-to-peer network, but not to make it truly possible for your mobile device or browser to be one of those nodes.

With the shift to mobile, we now live firmly in a world of clients and servers - with the former completely unable to act as the latter - and these issues seem more important to me than ever before. Meanwhile, ethereum actually refers to servers as "clients," so there isn't even a word for the actual untrusted client/server interface that must exist somewhere, and no one acknowledges that there will eventually be billions more clients than servers if it succeeds.

For example, whether running on a mobile device or on the Web, a dApp like Autonomous Art or First Derivative would need to interact with the blockchain in some way - in order to modify or render state (collectively produced artwork, its editorial history, NFT derivatives, etc.). However, this is practically impossible to achieve from the client side, as the blockchain cannot exist on your mobile device (or indeed in your desktop browser). Therefore, the only option is to interact with the blockchain via a node running remotely on a server somewhere.

A server! But, as we all know, people don't want to run their own servers. As it happens, companies have emerged that sell API access to Ether nodes running as a service, while providing analytics, enhanced APIs they build on top of the default Ether API, and access to historical transactions. Sounds familiar to ....... At this point, there are basically two companies. Almost all dApps use Infura or Alchemy to interact with the blockchain. In fact, even if you connect a wallet like MetaMask to a dApp, and the dApp interacts with the blockchain through your wallet, MetaMask is just calling Infura!

These client-side APIs don't use anything to verify the blockchain state or the authenticity of the response. The result is not even signed. An application like Autonomous Art says "Hey, what is the output of this view function on this smart contract", Alchemy or Infura responds with a JSON blob that says "This is the output " and then the application renders it.

This surprised me. A lot of work, effort, and time has gone into creating a de-trusted distributed consensus mechanism, but almost all customers who wish to access it do so by simply trusting the output of both companies without any further validation. This doesn't seem to be the best privacy scenario either. Imagine if every time you interacted with a website in Chrome, your request was first sent to Google and then routed to its destination and back. That's what's happening with Ether today. All write traffic is obviously already public on the blockchain, but these companies can also see almost all read requests from almost all users in almost all dApps.

Proponents of blockchain might argue that it's okay if these types of centralized platforms emerge because the state itself is available on the blockchain, so if these platforms misbehave, customers can simply move elsewhere. However, I would suggest that this is a very simplistic view of the dynamics that make platforms a reality.

Let me give you an example.

Making an NFT

I also wanted to create a more traditional NFT. most people think of images and digital art when they think of NFTs, but NFTs don't usually store this data on a chain. For most NFTs of most images, this is too expensive.

Instead of storing data on a chain, NFTs contain a URL that points to the data. What surprises me about these standards is that the data located at the URL has no hash commitment. Look at the many NFTs sold on the popular market for tens, hundreds or millions of dollars, and the URL usually just points to some VPS running Apache.

Anyone with access to that machine, anyone who buys the domain in the future, or anyone who destroys the machine can change the image, title, description, etc. of the NFT to whatever they want at any time (whether or not they "own" the token). nothing in the NFT specification tells you what the image Nothing in the NFT specification tells you what the image "should" be, or even allows you to confirm if something is the "right" image.

So as an experiment, I made an NFT that would serve a different image based on who was looking at it, since the web server serving the image could choose to serve a different image based on the requester's IP or user agent. For example, it looks one way on OpenSea and another way on Rarible, but when you buy it and view it from your crypto wallet, it always shows up as a big ? emoji. The NFT you bid on is not the one you get. There is nothing unusual about this NFT, that's how the NFT specification is built. Many of the highest priced NFTs can become ? emoji; I'm just making that clear.

A few days later, without any warning or explanation, the NFT I made was removed from OpenSea:

The deletion indicates that I am in violation of certain terms of service, but after reading the terms, I do not see anything that prohibits the NFT. The NFT changes depending on where it is viewed, which is how I describe it publicly.

However, what I found most interesting is that after OpenSea deleted my NFT, it also no longer appears in any crypto wallet on my device. This is web3, but, how is this possible?

Crypto wallets like MetaMask, Rainbow, etc. are "unmanaged" (private keys are kept on the client side), but they have the same problem as my dApp above: the wallet has to be running on a mobile device or in a browser. Meanwhile, Ether and other blockchains are designed with the idea that it is a peer-to-peer network, but are not designed in such a way that your mobile device or browser can actually be one of these nodes.

A wallet like MetaMask needs to do basic things like display your balance, your recent transactions, and your NFTs, as well as more complex things like building transactions, interacting with smart contracts, etc. In short, MetaMask needs to interact with the blockchain, but the blockchain is built so that a client like MetaMask cannot interact with it. So, like my dApp, MetaMask does this by making API calls to the three companies that are integrated in the space.

Again, just like my dApp, these responses are not authenticated in some way. They're not even signed so you can prove they're lying later. It reuses the same connections, TLS session tickets, etc. for all accounts in your wallet, so if you manage multiple accounts in your wallet to maintain some sort of identity separation, these companies know they are linked.

MetaMask doesn't really do much, it's just a view of the data provided by these centralized APIs. This is not a problem specific to MetaMask - what other options do they have? platforms like Rainbow are set up in exactly the same way.

What all this means is that if your NFT is removed from OpenSea, it will also disappear from your wallet. It doesn't matter functionally that my NFT is indelible on a blockchain somewhere, because the wallet just uses the OpenSea API to display the NFT, and it starts returning 304 No Content for querying owned NFTs by my address!

Reinventing the world

Given the history of web1 becoming web2, what I find strange about web3 is that technologies like ethereum have built many of the same implied pitfalls as web1. To make these technologies available, the space is consolidating around the ...... platform. Again. The people who will run the servers for you and iterate on the new features that emerge. infura, OpenSea, Coinbase, Etherscan.

Again, the web3 protocol is slow to evolve. When building derivatives, it is best to price minted derivatives as a percentage of the base value. This data is not on the chain, but in the API provided to you by OpenSea. People are excited about the way NFT royalties can benefit creators, but royalties are not specified in ERC-721 and it is too late to change it, so OpenSea has its own way of configuring the royalties that exist in the web2 space. Rapid iteration on centralized platforms has overtaken decentralized protocols and integrated control into the platform.

Given these dynamics, I don't think it's surprising that we're already in a place where what your crypto wallet thinks of your NFT is what OpenSea thinks of your NFT. I don't think we should be surprised that OpenSea is not a pure "view" that can be replaced, as it has been busy iterating on the platform beyond the strictly impossible/impossible to change criteria.

I think this is very similar to the email situation. I can run my own mail server, but it's functionally irrelevant to privacy, censorship resistance, or control - because GMail will be on the other end of every email I send or receive anyway. Once the decentralized ecosystem is centralized around a single platform for convenience, it becomes the worst of two worlds: centrally controlled, but still decentralized enough to be mired in time. I can build my own NFT marketplace, but if OpenSea mediates a view of all the NFTs in the wallets people use (and all the other apps in the ecosystem), it doesn't provide any additional control.

This is not a complaint about OpenSea, nor is it an indictment of what they are building. Quite the contrary, they are trying to build something that works. I think we should expect this kind of platform integration to happen and take into account the inevitable design systems that can give us what we want when things are organized this way. However, my feeling and concern is that the web3 community is expecting a different outcome than what we've already seen

In the Solana NFT space, there is no bigger player than Magic Eden. The marketplace, which launched last fall, typically accounts for 90 percent or more of all Solana transactions. It was valued at $1.6 billion in its latest round of venture capital funding in June of this year.

But with the rise of Magic Eden, members of the Solana NFT community - both creators and collectors - are increasingly concerned that the platform is becoming too centralized in its development. They point to recent updates that restrict access to third-party aggregators and tools, as well as the way Magic Eden hosts users' NFTs, potentially leaving users' assets vulnerable to attack. Marty, founder of Zion Labs, which makes the Solana NFT tool, told Decrypt: "People should be aware that hackers can get the keys to Magic Eden and 'rug' each of their NFTs. if it's decentralized and the code is open source, this won't be a problem. If the code was open source, this wouldn't have happened."

In its response to Decrypt, Magic Eden did not specifically mention the risks of the escrow-based transaction model, but it said it believes the alternative is currently less secure for users. The marketplace plans to adopt a no-escrow system in the future, but "the technology is not secure enough."

Magic Eden NFT Escrow Model Questioned

The debate is heating up over Magic Eden's practice of keeping users' listed NFT assets in an escrow wallet. instead of allowing them to remain in users' own wallets, Magic Eden escrows all live assets, and users' NFTs are kept in the escrow wallet via marketplace smart contracts. This approach was common in the early days of the Solana NFT market, but later entrants to the Solana ecosystem, such as OpenSea and Hyperspace, did not take this approach.

Last Wednesday, OpenSea tweeted "No to Solana Marketplace hosted NFTs" without directly naming Magic Eden, but the goal was clear. marketplace would limit choice and utility and jeopardize security."

Metaplex's auction agreement enables Solana to trade NFTs without the need for a marketplace to host the assets. A Metaplex source, who wished to remain anonymous, confirmed to Decrypt that Magic Eden's marketplace contract is based on an earlier version of the auction house, a license-free peer-to-peer trading system. However, Magic Eden has made significant changes to that contract code, as well as to the launch platform contract based on Metaplex's Candy Machine's Mint tool. magic Eden has also isolated them from the rest of the community. The source said, "They are closed source and licensed derivatives of the open source technology provided by Metaplex."

This approach increases the potential risk for NFT traders. Closed-source software cannot be audited by the community or benefit from a vulnerability reward program. Not even Metaplex knows what's in Magic Eden's marketplace contract code. What would happen if Magic Eden's hosted wallet was stolen? Or what would happen if Magic Eden suddenly went down, as some other crypto companies have done in recent months during the recent market crash, Metaplex sources said, adding that as of last week, the "centralized" hosted wallet held about 180,000 NFTs.

In response to a question from Decrypt, Magic Eden co-founder and CTO Sidney Zhang said the market plans to transition to an uncustodial model at some point, but in his team's view, the current solution is not secure enough. We are actively exploring the unhosted model and plan to move to it, but we believe that the smart contracts that other markets currently use to implement the unhosted model are not secure," he wrote. This shift raises many security issues and we want to proceed with caution to ensure that our users do not inadvertently lose assets by not having a timely updated list."

Magic Eden's many recent tweaks

In addition to its hosting model, Magic Eden has seen many new changes: increasing scrutiny, review of the way its platform operates, and the way third-party applications are built on top of it.

Last week, as the topic of Magic Eden's hosting model went viral on Twitter thanks to user "Pland": Magic Eden is "no longer a license-free Dapp" due to recent smart contract changes. Most users don't notice, but it does have a big impact on the ecosystem. Smart contracts hold the code that powers Dapp and NFT assets. Developers who spoke with Decrypt said that the contract change makes it mandatory for Magic Eden to sign every transaction that occurs on its marketplace, unlike before. As a result, a number of third-party applications that aggregate multiple marketplace listings and so-called "sniper bots" that can be used to buy specific NFTs have been compromised.

Magic Eden acknowledged the contract change to Decrypt, explaining that transactions now require two signatures: one from the end user and another from an API key provided by Magic Eden. the API key is used to authenticate developers and third-party programs that wish to access the application or service. Ether-centric marketplaces like OpenSea also have API systems.

Zhuojie Zhou, co-founder and chief engineering officer of Magic Eden, told Decrypt: "This change was introduced to maintain the reliability of the core site and reduce bot behavior that could compromise user listings and transactions. We are very open to ecosystem participation in our API program." Solana Labs has recently made some changes in an attempt to improve the stability of the network.

Zhou said Magic Eden has made more than 300 API keys available to developers, including aggregators like Tensor and NFT Soloist, as well as wallet app developers like Exodus and Slope. He also noted that developers of the Solana wallet Phantom have asked Magic Eden to have an API to verify that transactions are coming from their servers. "We believe in supporting a formal developer ecosystem to achieve our goal of a safe and secure marketplace," Zhou added, "and we keep an open mind to develop the API based on the needs of our partner developers."

Magic Eden's Mandatory "Anti-Insideration Initiative"

However, some developers in the Solana space see the shift as a rejection of decentralization principles. a representative from Hyperspace, the NFT marketplace aggregator, told Decrypt, "We're surprised they did this because it's completely centralized and doesn't benefit the end user. Because it increases the dependency on their servers, which leads to an increase in transaction failure rates."

Before the contract change, Magic Eden contacted Hyperspace and threatened to "shut down Hyperspace if they didn't change their platform to serve them," the person, who asked not to be named, said, adding that Magic Eden allegedly wanted Hyperspace allegedly wanted to provide Magic Eden with "exclusive uptime information and to operate only through their API.

A Magic Eden representative denied threatening them in the discussion: "We encourage our partners to integrate with Magic Eden as deeply as possible in order to provide the fullest possible technical and operational support. Unfortunately, Hyperspace is not interested in such a partnership and has been hostile."

Hyperspace says it found a solution to the Magic Eden API and continues to offer aggregated lists, but other aggregators, such as CoralCube, apparently lost functionality as a result. "Since then, they have been trying and actively working on ways to stop us," a Hyperspace representative claimed.

"CoralCube used to have a migrate list button, but Magic Eden recently moved to Web2's centralized platform. Items cannot be removed without Magic Eden's centralized signature. That's why we removed the migration button and now NFT is stuck in Magic Eden's hosting."

Some of Solana's builders told Decrypt that they believe Magic Eden's move was intentional to exclude the NFT aggregators that have gained attention in recent months. Hyperspace has been speaking out against this strictly anti-competitive behavior because it violates the principles of the open web.

Magic Eden's new features have come under intense fire

In addition, Magic Eden has come under fire for implementing new features that appear to have been inspired by external Solana applications. Last week, Magic Eden's go-live feature, which allows projects to create user permission lists before NFT deletes them, was resisted for being very similar to Blocksmith Labs' Mercurytool.

Anonymous NFT collector Topo Gigio told Decrypt about Magic Eden's new addition: "It seems like an unwarranted attempt to exclude anyone who could do better." Meanwhile, Zion Labs' Marty claims, "Magic Eden is 'using venture capital as a weapon' to rapidly expand into an all-in-one Solana NFT resource."

Magic Eden's Zhou responded that Magic Eden is a "user-first company" that adds features based primarily on user requests. He claimed that the extensions on the platform are for NFT collectors and rejected the debate about centralization, saying, "This conversation is not about centralization and decentralization, it never has been. Collaborative tools have existed since our inception, based on Magic Eden's evolving market experience, and we don't intend to change that."

For some players in the Web3 space, the overall conversation around Magic Eden has been largely about centralization versus decentralization, including how key players in the space should approach issues such as asset escrow, open source code, and composability of blockchain assets and protocols. Between continuing to use third-party escrow and API-centric changes, Magic Eden's decision recently hasn't been for everyone. But Magic Eden remains the best choice for Solana collectors to buy and sell NFTs.

Criticism of Magic Eden is growing, but it remains to be seen whether many NFT projects will choose to distribute elsewhere, and whether prominent collectors will choose to take a public stand and exit the Magic Eden market. The NFT collector tweeted that he was giving up liquidity and claimed he would no longer use the market, noting that Magic Eden's escrow policies and contracts had changed. "All the liquidity is at Magic Eden and my exit is irrelevant to them," he told Decrypt, "I'm happy to move my high value assets elsewhere, even if the volume is smaller."

As we all know, DAOs do not have the desired level of autonomy and automation. They are implemented more as simulations of the vision of a "fluid on-chain governance process", whereas we strive for a more dynamic and intelligent type of DAO.

In terms of granting membership and voting rights to users, today's DAOs use one of two approaches.

Token-based governance Users gain membership by holding, holding in trust, pledging or earning tokens

Social-based governance Users gain membership by gaining sponsorship, contributing to a DAO, or signing up

These mechanisms vary in terms of the level of permission not required for membership, voting rights, and participation thresholds. For example, one needs to hold DeveloperDAO tokens to become a member, or 10 million UNI to initiate a proposal to the DAO.

How dynamic can we get?

Some DAOs use tools like CollabLand to create token taps on discord to reward members for outstanding contributions, or Coordinape, which allows members of the community to review each other's contribution levels. These are all great ways to get contributors more voting power, but none of them work very well in terms of adjusting or removing voting power. Our vision for the protocol also relies on evaluating and recognizing user contributions in some autonomous way, which will become a future work in progress.

Membership in a DAO should be fluid. We need to start designing the DAO in a way that captures the dynamism of web3 participants, but also recognizes that it is cheap for these participants to switch between working on different protocols.

The question that needs to be thought about is.

How can we reduce voting rights, or expel someone from the DAO?

How can you set behavioral thresholds for DAO members in the protocol, not just voting rights?

How can membership in a DAO be changed over time?

How can we ensure that DAO participants-who have a significant impact on the governance agreement, the culture, and the longevity of the DAO-are still the best people to make decisions for the benefit of the DAO?

Being a member of the DAO should be a privilege, not a right.

Anyone can buy protocol tokens, attend town hall meetings, or join the discord channel. But just because they hold a token, or change their avatar to that of a cat, does not give the DAO the right to govern or manage the vault (extremely important!) The right to govern or manage the vault (extremely important!) cannot be given to them forever and 100%.

The only DAO mechanism (that I know of) that removes members from the chain is MolochDAOv2's guildkick. It allows members to vote to expel other members, usually against those who have acted maliciously. This mechanism is often used as a preemptive defense against the potential risk to the DAO from angry withdrawals and divestment. However, such a highly hostile initiative does not meet our requirements for liquidity within the DAO. The guildkick remains a static proposal that can only be submitted at the right time, rather than an autonomous solution.

Reputation-based DAO

Holding tokens can indicate a person's interest in being part of it, but simply holding tokens is not enough to demonstrate sufficient commitment to the protocol and the well-being of the DAO. It is the user's interaction with the protocol and the community that really matters for DAO membership and voting rights - it is reputation.

DAOs can measure dedication in their protocols based on network participation, governance participation, and community participation. a DAO cannot rely on token ownership alone, but should require its members to participate in the protocol and the DAO to build reputation.

If whale investors don't want to pledge their tokens, should they be given very large governance rights? If they don't even have a protocol POAP, can they still be allowed to sit down together and work together on big ideas?

Membership in a DAO should be fluid.

For token-based voting, the switching cost for users is low - a few simple actions on Uniswap and you can buy and sell any token and join any DAO, but DAOs can't currently handle the same level of liquidity.

DAOs need membership to be scalable - users' membership should be based on their ability to consistently meet reputation metrics, not static one-off activities or purchases. If they can't meet the minimum requirements on the chain, they should no longer be part of the DAO. Imagine how much better organized a DAO would be if the membership list could be automatically updated, always leaving the highest quality, most engaged members. It would be like a country in which only citizens have the right to vote.

Let's simulate.

Alice joins BobbyDAO, which oversees the Bobert protocol. currently Alice is a whale holding $BOB. Because she meets the $BOB holding threshold, she can submit proposals and easily sway votes. However, Alice has neither pledged $BOB in the agreement nor responded to a single community call! So Alice shouldn't have those rights.

Recently Bob has been thinking about strengthening his DAO. he feels that the reputation mechanism should work. That way, more $BOB pledgers would have a voice in the agreement than $BOB holders who don't really care about the best interests of the agreement. He made the decision that all users who have pledged $BOB for at least 30 days in the past year will be allowed to retain their membership. At any time, if a user fails to meet the 30/365 ratio criterion, they lose their membership. If that user repledges, then DAO welcomes him back. In this way, Bob hopes to keep the best protocol contributors within the DAO.

The future of the DAO is a community that can be maintained programmatically and is constantly changing. Members are the lifeblood of the DAO, allowing the DAO to focus on its mission while benefiting from it at the same time. A DAO membership that fails to organically cultivate a reputation within the protocol is perhaps the wrong arrangement and can lead the community towards jingoism or authoritarianism.

Users will buy many tokens, join many discord, try and leave many DAOs. a few will stay and will use their tokens for the protocol, and even fewer will become long-term contributors. The success of a DAO is not measured by how many members it has, but by the quality of its proposals and its ability to keep contributors in the protocol (or whatever else this DAO manages). A member that is useful today may not be useful tomorrow. We need to learn to build the DAO with soldiers of the stream, which is why it is so important to keep the DAO dynamic, and allow membership to evolve.

Fortunately, most of the interesting reputation metrics are on-chain or accessible through the prophecy machine, and are available for DAOs to adopt in their governance systems.