CVS, which owns health insurer Aetna, on Wednesday slashed its full-year profit outlook, citing the potential for higher medical costs to bite into its profits. That warning came two weeks after insurance giant Humana cited the same factor as it issued a dismal 2024 earnings guidance.

Medical costs from Medicare Advantage patients have spiked over the last year as more older adults return to hospitals to undergo procedures they had delayed during the Covid pandemic, such as joint and hip replacements.

Medicare Advantage, a type of privately run health insurance plan contracted by Medicare, has long been a key source of growth and profits for the insurance industry. More than half of Medicare beneficiaries are enrolled in such plans, enticed by lower monthly premiums and extra benefits not covered by traditional Medicare, according to health policy research firm KFF.

But investors have become more concerned about the runaway costs, which insurance companies say may not come down anytime soon. Other companies in the Medicare Advantage space are UnitedHealth Group and Elevance Health.

CVS executives said on an earnings call Wednesday that the company’s insurance division saw slightly higher rates of outpatient care, including hip and knee surgeries, in the fourth quarter. They also saw more use of supplemental benefits such as dental and vision care, and “some pressure” from RSV vaccinations.

The executives said inpatient care, or formal hospital admissions, was in line with the company’s expectations for the period.

The insurance segment’s medical benefit ratio — a measure of total medical expenses paid relative to premiums collected — increased to 88.5% for the fourth quarter from 85.8% during the year-ago period. A lower ratio typically indicates that the company collected more in premiums than it paid out in benefits, resulting in higher profitability.

Last month, Humana said it saw an even bigger jump in medical costs in the fourth quarter. The company said the increase came partly from higher outpatient activity, but the company largely blamed it on an unexpected increase in inpatient care in November and December.

That pushed its medical benefit ratio in its insurance segment to a whopping 91.4% for the quarter, up from 87.4% for the same period a year ago.

Higher medical costs may be a larger problem for Humana than they are for CVS and other insurers. That’s because Humana is more dependent on its Medicare Advantage business than its rivals, as it accounts for more than 80% of its earnings, UBS analysts said in a Jan. 25 note.

They added that there is no other part of Humana’s business that could meaningfully dampen the hit from higher medical costs on the insurance side. Humana has a specialty pharmacy segment called CenterWell, but it only brought in roughly a fifth of the revenue that the company’s insurance division booked for the fourth quarter.

Meanwhile, CVS has a retail pharmacy business and a health services segment, both of which posted stronger-than-expected revenue for the quarter.

Another insurance giant that has been seeing higher medical costs, UnitedHealth Group, also has large health-care services and pharmacy operations that diversify its earnings streams.

The bigger question for all three companies is how exactly a new policy called the “two-midnight rule” will impact their insurance businesses.

Starting this year, Medicare Advantage plans have to cover their members’ hospitalizations at the higher inpatient rate if their doctors predict they’ll have to stay beyond two midnights. That policy has applied to traditional Medicare plans for nearly a decade.

If not heard personally, no one would have expected this passage from the mouth of "Sister Lin". And its "conversation", is four dynasties back to the "poet" Li Bai.

At the beginning of May, B station up master "AI-Talk" released the work "AI Li Bai and Lin Daiyu poetry". 3 minutes and a half of the video, every word is generated by GPT4, similar "mix and match" video and "AI Nietzsche to talk about AI clowns" "Tom Cruise dialog Kant" and so on. AI Nietzsche on AI Clown", "Tom Cruise Dialogues with Kant" and so on.

"The content is harmonious, the language is exquisite, and the logic is good. It can be seen that the AI's level of text creation has reached a medium level, 80 points upwards." Jin Yuanpu, director of Renmin University of China's Institute of Cultural and Creative Industries and professor at the School of Arts, commented.

Recently, Jin Yuanpu was a guest at the Shenzhen Citizen's Cultural Lecture Hall on the topic of "The current development of the big model of artificial intelligence and the new situation of the 100-model war". The traditional cognitive talk about poetry and song professor of the Faculty of Letters turned to talk about AI, you can see the big model in all walks of life "out of the circle" degree.

Jin Yuanpu said in an exclusive interview with a reporter from Southern Finance and Economics that ChatGPT, a type of AI language model, cannot be understood only from a technical point of view. Because the big model can be integrated with all the words, culture and thinking left behind in the past era, it is actually about the boundaries of human intelligence, and the exploration of the big model should focus on the long term.

In the short and medium term, as long as the big model is "fed" enough, it will be able to generate qualified content, and therefore become a double-edged sword - empowering creators and replacing or partially replacing them.

The use of AI is especially prevalent in the fields of literature, journalism, foreign languages, and fine arts. Eye-catching news headlines such as "30 seconds to generate a picture, the cost as low as 20 cents" and "AI machine flip to reduce the cost to 1%" have caused a large number of "workers" to be replaced by the occupation of anxiety. Anxiety.

Conveyed to the younger generation, with the admission scores of colleges and universities across the country one after another, many colleges and universities news, Chinese, small languages and other traditional liberal arts majors in varying degrees of "cold".

"The belief that literature, new communication, small languages and other related professions will immediately be replaced by AI is actually an exaggeration of the impact of AI." Jin Yuanpu believes that, with a large number of public materials as a reference, AI can help complete the repetitive and consumptive work at the practical level, and its work is often higher than that of the medium and low level industry practitioners. And once you delve into fully autonomous and innovative literary creation, including the expression of vivid and delicate human emotions and the exploration of deep philosophical issues, AI is not up to the task.

AI with "moderate to high" grades

In February this year, Jim Mullen, CEO of Reach, publisher of the Daily Mirror and Daily Express, said publicly that he would explore the use of ChatGPT to assist journalists in writing short reports on topics such as local weather and transportation. During the same period, it was rumored that a domestic financial media outlet issued a "Notice on the Use of ChatGPT by All Editorial Departments".

Gradually, enabling ChatGPT in news creation is no longer new, and practical tips for media people have begun to appear on the Internet, including how to quickly generate headlines, how to write TV or radio promotional scripts, how to quickly generate social media posts such as tweets and collaterals from published news content, and so on.

"The quality of AI generation depends on how well the big models are pre-fed. In some relatively simple areas, if the big model has already focused on more comprehensive background material, it can instantly produce a good quality story, at least moderately high, which will greatly enhance news industry productivity." Jin Yuanpu said.

Ditto in the literary field. Recently, the ReadWrite Group released the first large model of the domestic online literature industry, "ReadWrite Wonderful Brush," which is said to understand the content (story, characters, and worldview settings), creation (understanding the creative techniques of online literature), and the "online literature terriers" formed in the process of interaction between writers and readers. ".

According to the statistics of the Institute of Literature of the Chinese Academy of Social Sciences, China's online literature market in 2022 amounted to 38.93 billion yuan, an increase of 8.8% year-on-year, and the cumulative number of online literature writers exceeded 22.78 million.

In the face of such a huge market, Jin Yuanpu said that if writers can effectively leverage the big model, human-computer collaboration will realize the level of writing from "medium to high" to "excellent", and the time and labor costs will be significantly reduced.

However, AI is also a "double-edged sword", so that human beings on the one hand, pleased with the liberation of labor, and on the other hand, fell into the once "Lee Sedol defeated Alpha Dog" type of anxiety: whether AI will be an all-round "victory" over human beings? AI will "defeat" human beings in all aspects?

In March, the famous American linguist and philosopher Chomsky responded: "The human brain is an extremely efficient and even elegant system that requires only a small amount of information to operate; it seeks not to infer brute correlations between data points, but to create explanations."

In the comment section of the video "AI Li Bai and Lin Daiyu Poem Pairing", there are also many netizens who have found the pain point of AI creation, "There are rhymes but no chapters, and there are words but no feelings."

How humans and machines coexist

Even if human beings have their emotional perception that can not be replaced by ChatGPT, at the level of social opinion, the panic that liberal arts majors and jobs will be replaced by artificial intelligence is still rampant, and even affects the enrollment of this year's college entrance exams for traditionally liberal arts majors such as journalism, Chinese language, and small languages at some domestic colleges and universities.

In the face of AI this "double-edged sword", man and machine how to explore a mutually reinforcing state of balance?

Jin Yuanpu believes that the "useless theory of liberal arts" actually exaggerated the influence of AI, but also underestimated the subjective initiative of human beings. But in the age of science and technology, the training mode of "liberal arts students" does need to be optimized and adjusted, focusing on improving its cross-border composite ability.

"Liberal arts talents should have at least threefold composite nature. One is to have received the cultivation of literature and art, literature and art is a hotbed of creativity; the second is to master a certain amount of technical knowledge, to be able to use the latest scientific and technological achievements to improve the quality and efficiency of learning; the third is to master a certain amount of knowledge of finance and accounting, with the ability to manage." Kim Wonpo explained.

Just with other business half-yearly test is different, such as new energy vehicles, cell phones, e-commerce platforms and other business forms of the half-yearly test, there is enough public data and information to support, easy to analyze the evidence, while the big model is still in a "black box" state, did not run out of a clear business model, the so-called data and information and other arguments are also The so-called data information and other arguments are also unavailable.

It is quite a joke that even from the perspective of product function, the big model has never given birth to a universal evaluation tool. For the ultimate goal of AGI, there are naturally a variety of evaluation methods, such as the "Squirrel and Mandarin Fish Method", which is the standard for C-suite users in China to "evaluate" big models.

For this reason, most of the major domestic manufacturers have not been able to open up the use of their own large models like OpenAI, but have instead implemented internal testing mechanisms.

Big model more landing exploration to the B end and G end tilt, for example, Tencent preemptive industry big model, as well as Huawei's Pangu 3.0, Jingdong Rhinoceros and so on. As the current head players focus on the track, its big model favors to show as much as possible the mature product form, with commercialization landing as the basic goal. For example, in order to rapidly promote commercialization of this type of large model downward popularity, in addition to the business landing-oriented, the localization and deployment capabilities have also become important reference indicators.

Even so, in the view of industry insiders has been "sent to the bowl in front of the" industry model is still a lack of enterprises to buy, industry model of the wind since June has blown a month, so far there has not been a large-scale commercial cooperation.

Therefore, it is not difficult to see that in today's investment market, the investment related to large models is concentrated in the secondary market rather than the primary market. Even if Wang Huiwen this level of bull entry, public news said its A round of financing is much higher than 230 million U.S. dollars, and its financing ability compared to Microsoft from time to time to receive tens of billions of U.S. dollars to feed the OpenAI is not comparable.

The investment market is a qualified barometer. Obviously, the domestic big model in the half-year examination of the time node submitted by the answer sheet is not satisfactory, but also need a period of dormancy and polishing, in order to let the "story" come true.

Big model has no business model?

The business model should be at the forefront of the need for the country's big models to respond to the market's skepticism.

ChatGPT this has long occupied the user's mind of the head of the chair there is a significant decline in heat, the earliest release of the domestic general large model Baidu and Ali two also in a number of players to follow up after the "silence". The reason for this is that the business model of the generic large model failed to run through. Even in the court of public opinion has been recognized by users, but the commercial closed loop has never appeared.

To test the wider range of Baidu big model, for example, its commercialization application Wenxin Qianfan payment model is to call the number of token generated charges, the standard is 0.012 yuan / thousand tokens, the output of a thousand words of text to spend 0.12 yuan.

Leaving aside the speed of its cost recovery, 0.012 yuan / thousand tokens of charges seem cheap, but text generation often requires multiple interactions to obtain the desired results, multiple interactions prompt will increase the hidden cost of unlimited, after all, Wenxin Chifan is not a wave of employees.

Similar to the scene is the question and answer community, academic Sun Quan (a pseudonym) told Photon Planet, the use of the model application experience is similar to the search for high-quality answers in the question and answer community, its user thinking is the problem of granularity, and the willingness to pay is often only generated after finding high-quality answers. Therefore, Baidu has chosen to reason the number of texts as a payment criterion, only at present it is not possible to cover the commercial hidden costs.

If the B-side of the popular monthly payment, that is only the cost of the party from the user to their own, obviously not a long-term solution. ChatGPT face C-side users under the pricing of $ 20 / month, there is still a suspicion of jerry-building is the best evidence.

At present, the commercialization of general large models, whether B-end or C-end, is difficult to achieve break-even, but also likely to encounter compliance risks such as AI ethics, regulation and so on. Therefore, the industrialization and verticalization of large models have become a paradigm shift under the demand for landing.

On the contrary, although the industry model, although its product form began to land demand, but in the actual landing of the problem has yet to be solved.

A kind of reference case is the vertical to C model built on the basis of its own product ecosystem, for example, Zhihu announced early in the product of the internal testing of Zhihaitu AI and Ctrip asked released not long ago.

The advantages of these two companies in entering the large model track are the same, which lie in their own community ecosystems and the high-quality community content derived from these ecosystems. The content, as industry data, can become the training corpus for big models after simple cleaning. The subtle difference between the two is that Zhihu has been a content community since its inception, while Ctrip has only begun to make efforts to do content in recent years.

However, both Zhihu and Ctrip seem to have failed to address the user's pain points and improve their existing functions in the form of big models.

Zhihaitu AI has announced its product "Hot List Summary", which is to capture high-quality Q&A through AI and rewrite the synopsis to present to the user, while another application "Search Aggregation" is to aggregate views from the answers to improve the efficiency of users in obtaining information and forming decisions. efficiency of users' access to information and decision-making.

Recommendation, hot list, a kind of aggregation function is Zhihu "traditional arts", the performance of the big model empowerment in the user level did not set off a splash. Moreover, the process of AI rewriting and embellishment also covers the personalized features of popular answers, and for users, the function of this application only lies in the quick understanding of information, which is contrary to the differentiation and personalized communication advocated by the content community.

The OTA-based Ctrip Ask, in the opinion of Ctrip's Chairman of the Board, Liang Jianzhang, is a "reliable answer bank" for the tourism industry. The effectiveness of its products need time to test, but since the positioning of the view, there is the same "to the end" suspicion.

Tourism in the eyes of young users in the eyes of there is no standard answer, "special forces", "punch card", "immersion" and other diversified forms of tourism has proved this point. If we assume that a large number of users make travel route plans through AI, the uniform route plans will affect the community communication and atmosphere, and even lead to a decline in the user's stay time.

Generally speaking, it seems that the attempt to realize vertical modeling in the C-suite is not smooth, and even has the possibility of becoming a "sunk cost". Perhaps influenced by the myth of "improving efficiency", the product positioning is mostly limited to the word "efficiency", which is not a core dimension of user experience.

The same paradigm has been demonstrated in the field of to B, and in the pursuit of efficiency of the B-side, the business model of the industry big model and the implementation of the problem has been more profoundly demonstrated.

The black box that can't be understood

"AI is not physics, there are rarely any theoretical major technological breakthroughs, it is more about fine-tuning and small optimizations in dimensions such as model structure, data quality, etc., and even many times the model output is better and the team can't find out why."

In the view of an industry insider, the big model in the industry and outside there is a huge cognitive bias, and the reason is that the big model training and the AI industry for the outside world is a no compromise "black box", it is difficult to scrutinize the big model to produce the output results of the reasoning process, which can not be seen and cannot be touched.

This leads to the outside world in the ChatGPT to bring the frenzy period, once calmed down, will be on the big model of the "black box" attitude of caution. This will lead to the big model in the landing of the dilemma, and this phenomenon is more obvious in today's to B route to change the process.

To today's clear to B route of the big factory products, for example, including Tencent Cloud launched MaaS technology solutions, Huawei Cloud launched Pangu big model, relying on its own cloud computing ecosystem, are said to support the deployment of its big model services diversified deployment, including cloud deployment, localization and rapid deployment and so on. There are also achievements in interaction, operation, and subsequent addition of new industry data iterative optimization, etc. It can be said that the threshold of the big model has been reduced to a very low level in order to land.

However, the cognitive wall brought about by "prudence" has not been broken, and even though the wind of ChatGPT has been blowing for half a year, many enterprises do not have the motivation or interest to study how to import large models.

A few years ago, the cloud computing industry can be seen following a similar logic. Cloud computing is in the recognition of the value of data, as a basis for services and derivatives, as for the value of the big model in the enterprise, relatively speaking, it is a leap in the value of data. The same is the lack of technical capabilities of enterprise customers, even the popularization of cloud computing in the domestic enterprise is still far from the end of the road, the big model needless to say.

Industry model is good to use or not, in fact, has been unimportant, after all, the use of commodity value ultimately need to be tapped by the user. What's more, outsiders will roughly measure the level of the model through certain tests and performances, such as the "Squirrel and Mandarin Fish Method" or the Huawei Pangu Weather Model, which has been challenged recently because of errors in predicting the landfall location and intensity of the mega typhoon "Dusu Rui".

Perhaps this is why the recently released Jingdong Rhinoceros Big Model has chosen to prioritize its own business scenarios and is expected to open up to "external serious business scenarios" early next year.

What is worth mentioning is that, "the industry into the wind" under the commercialization of the so-called industry-oriented model to replace the original big model of the "general" narrative at the same time, but also suffered a lot of people's "lost". "The

The definition of the so-called industry model is ambiguous. A Foundation Model is not about the number of participants, but about the generic capabilities that emerge from training with generic data. If the same model architecture is used, but a single domain data is used for the data, not only the generic capability is lost, but even the domain problem cannot be solved due to the discount of emergence.

If the use of industry data on the basis of the original model to do the second pre-training, the equivalent of fine-tuning the original model, then the product itself is still in the model layer, can be called the industry model; such as through the Prompt or plug-in database to join the field of knowledge, it is only on the original model to stimulate the ability of the product should be attributed to the model of the application layer above the industry model is overstating the case.

At present, the vast majority of large manufacturers in the development of industry model are the former, such as Tencent, Jingdong, Huawei and so on. The latter is due to lighter investment and rapid improvement in the performance of the model ability, more will appear in the open source community, such as a period of time ago sparked a heated debate on the legal big model ChatLaw.

"Compared with the former, the latter is more mature in product form, which facilitates the rapid construction of modeling capabilities, but the latter tends to have a higher ceiling after completing the process of instilling domain knowledge", said an industry insider.

Open source threat

Recently, Meta made its latest open source big model, Llama2, freely available under an open commercial license and introduced it to Microsoft's Azure platform, a move that has been hailed as a major milestone for open source LLM and has even begun to threaten the position of closed source head honcho OpenAI.

Through Microsoft, the big model gold owner, Meta challenges OpenAI with a more open stance.

In fact, the "open source faction" has quietly risen as a third party before. "We don't have a moat, and neither does OpenAI." This sentence came from an internal document accidentally leaked by Google in May. Its content is to the effect that on the surface, OpenAI and Google in the big model you catch up with me, but the real winner may not come from the two, the reason for this judgment lies in the increasingly rich open source ecosystem.

The open source ecosystem has become more and more active, and even the emergence of Llama2, a representative of the modeling ability, and LORA, a representative technology of the Finetune (model fine-tuning) paradigm, all of which have made the closed-source giant vendors striving for "vigorously producing miracles" feel a clear chill.

Open source technology sharing and talent flow and other factors, but also in the big model of the black box more and more "glass", the lack of barriers to the inevitable result is a large factory in a huge amount of money, time investment in the Konw How easy for the open source community to be overthrown.

Domestic head of the big manufacturers to respond to most of the "two-handed approach". The left hand "shut the door to build a car", in the form of small-scale internal testing to continuously polish the product form and capabilities, the right hand "brainstorming", based on the cloud developer ecosystem to create an open source community within the ecosystem, but this just requires vendors from the computing power layer, model layer to the application layer of the full stack of the layout. Layout. Ali cloud launched a large model open source community magic ride GPT, Huawei cloud, Baidu cloud, Tencent cloud also have layout.

Overall, whether it is the industry or general, to C or to B, the big model of the six-month test gives us a direct feeling: landing difficulties, profitability expectations continue to move back; the risk of gradual strengthening, it is difficult to say that the technical barriers. So, where is the road to break the current situation?

For now, there are two interesting directions. One is the vector database known as the "Memory of the AI era", and the other is the intelligent hardware empowered by model intelligence.

The so-called vectors are multidimensional data that can represent anything, including text, which is most valued for LLM training today, as well as images, videos, audio and sound. These forms of content are clearly represented in a database and support semantic retrieval, i.e., retrieval by similarity, e.g., man vs. boy. In other words, vector retrieval is the SEO of large models.

As mentioned above, domain knowledge can be fine-tuned or externally tuned to improve the construction and use of industry models through vector database capabilities, which is naturally where the next phase of power lies for the big players. Since May, capital has been pouring into vector data-related tracks, and as an application layer product with a more certain outlook, vector data has also gained the close attention of a number of VCs.

As for the built-in model of intelligent hardware, it is a leap in the ability of intelligent assistants such as "siri" and "Ai", as well as an outreach to real intelligent devices (cell phones and computers). The open source community has long been the big parameter model built-in MAC attempts, while the big manufacturers in the past mobile Internet era has accumulated a certain amount of hardware production capacity, relatively speaking, its first-mover advantage is more obvious.

Less PR-style spring and autumn brushwork, landing has become the core needs of the big model is no longer mysterious, the story is also less and less, began to "deep dive" track players are still making efforts. The industry needs the next "ChatGPT" moment, so that we can see the divers surface and confront each other head-on.

Marc Whitten, president of Unity Create, dedicated a post last month to the profound impact that generative AI will have on Unity. They believe that the power of generative AI will make Unity creators much more productive, while ushering in a flood of new creators who will face lower barriers when building RT3D games and experiences. The current generative AI Cambrian explosion will be an opportunity for Unity to grow further.

Unity Muse

Unity Muse is an AI platform that accelerates the creation of real-time 3D applications and experiences such as video games and digital twins. the ultimate goal of Muse is to enable you to create virtually any content in the Unity editor using natural inputs such as text prompts and sketches.

Unity also announced that it has begun closed beta testing of a Unity Muse platform product, Muse Chat, which is a key feature of the Muse platform. Muse Chat helps you find relevant information, including working code examples, to speed development and solve problems.

More features will be created next, including the ability to create textures and sprites and even fully animated characters, all using natural input.

Unity Sentis

Unity Sentis is also in internal beta, and Unity sees Sentis as a true game changer.

The biggest highlight: combining neural network technology with Unity.

On a technical level, it connects neural networks to the Unity runtime, unlocking huge possibilities for Unity. Simply put, Unity Sentis enables AI models to be embedded in the Unity runtime of a game or application to enhance gameplay and other features directly on the end-user platform.

Sentis allows AI models to run on any device running Unity.

As the world's largest game engine, more than half of all games across all platforms (PC, console and mobile devices) are created using Unity, including mainstream games with billions of users such as King of Glory and ProtoGod. It is conceivable that King will have more elements of AI-generated content in the future.

Unity says Sentis is the first cross-platform solution to embed AI models into a real-time 3D engine, building and embedding your model once to make it run on multiple platforms, from mobile devices to PCs, from the Web to consoles like Nintendo Switch and Sony PS. And because it runs on the user's local device, there is also none of the complexity and latency associated with cloud hosting.

AI Hub drives Unity shares soaring as investors get bullish

In addition to its two AI products, Unity also announced the launch of AI Hub, an AI software marketplace for developers, providing a more convenient AI software trading platform for AI software developers and game developers. Users can find curated solutions for AI-powered development and game play.

Unity will allow AI software developers to supply development software directly to game developers through the AI Hub, in third-party packages that meet Unity's highest quality and compatibility standards. Unity will also charge an intermediary fee.

You can find professional-quality AI solutions from providers such as Atlas, Convai, Inworld AI, Layer AI, Leonardo Ai, LMNT, Modl.ai, Polyhive, Replica Studios, and Zibra AI. These new solutions help support and accelerate your creative process by supporting AI-powered intelligent NPCs, AI-generated visual effects, textures, 2D and 3D models, generated speech, AI in-game testing, and more.

Investors were also very realistic as Unity stepped onto the AI bandwagon with two AI products and the AI Hub, with shares closing up 15% yesterday.

This is not the American "Netflix Festival" VidCon, but a scene from the developer conference of Snowflake, a famous American data platform.

On June 26, local time, Nvidia founder Jen-Hsun Huang and Snowflake CEO Frank Slootman discussed "how to bring generative AI to enterprise users". The moderator was the former GP of Greylock and now founder of investment firm Conviction.

At the meeting, the "Godfather of Leather" was as surprising as ever, compared to the professional managerial sophistication of the "host" Frank, who not only called the collaboration "We are Lovers, not Fighters", but also joked that the trained models provided to Snowflake were equivalent to a "10% discount" for customers.

Snowflake users can directly use NVIDIA's pre-trained AI models to analyze their own company's data on the cloud platform and develop "AI applications" for their own data, without leaving the platform.

"The major change now comes from data + AI algorithms + compute engines. Through our partnership, we're able to bring all three together," said Jen-Hsun Huang. said Jen-Hsun Huang.

Talking Points:

Big language models + enterprise-specific databases = problem-specific AI applications; What used to be Data going to Work is now Work going to Data, allowing computation to go where the data is and avoiding data silos; NVIDIA provides pre-trained models that have been trained in the NVIDIA AI factory at a cost of tens of millions of dollars, so calling the compute engine on Snowflake is already "0.5 percent off"; Software 3.0 era, based on models, databases, enterprises are able to build their own proprietary applications in a matter of days; a future where enterprises are able to produce many intelligent agents and run them; The real challenge for enterprises is the mixed structured, unstructured data, how to be mobilized. This may be able to bring about a renewal of business models.

The following is the main content of the conversation between the two parties, as compiled and edited by Geek Park:

01 Talking about cooperation: Bringing the best computing engine to the most valuable data

Frank:

NVIDIA is currently playing an important role in history. For us, the ability to bring data and large enterprise relationships. We need to enable this technology, as well as get the entire service stack to use it effectively. I don't want to use the term 'match made in heaven,' but it's a great opportunity for a layman to get in the door of that opportunity.

Jen-Hsun Huang:

We are lovers, not adversaries. We want to bring the best computing engine in the world to the most valuable data in the world. Looking back, I've been working for a long time, but I'm not that old. frank, you're a little older (laughs).

These days, for reasons that are well known, data is huge, data is valuable. It has to be secure. Moving data is difficult, the gravitational pull of data is real. So it's much easier for us to bring our compute engine to Snowflake. Our partnership is about accelerating Snowflake, but it's also about bringing artificial intelligence to Snowflake.

At its core, it's a combination of data + AI algorithms + compute engine, and our partnership brings all three of those things together. Incredibly valuable data, incredibly great AI, incredibly great compute engines.

What we can do together is help our customers use their proprietary data and use it to write AI applications. You know, the big breakthrough here is that for the first time, you can develop a large language model. You put it in front of your data, and then you talk to your data as if you were talking to a person, and that data will be augmented into a large language model.

The combination of a large language model plus a knowledge base equals an artificial intelligence application. This is simple, a large language model turns any data knowledge base into an application.

Think of all the amazing applications that people have written. At its core is always some valuable data. Now you have a query engine generic query engine up front that's super smart and you can make it respond to you, but you can also connect it to a proxy, which is the breakthrough that Langchain and vector databases have brought. The breakthrough stuff of overlaying data and big language models is happening everywhere, and everybody wants to do it. And Frank and I are going to help everyone do that.

02 Software 3.0: Building AI applications that solve a specific problem

Moderator:

Looking at this change as an investor, Software 1.0 was very deterministic code, written by engineers according to function; Software 2.0 is optimizing a neural network with carefully collected labeled training data.

You're helping people pivot to Software 3.0, which is a set of underlying models that have incredible capabilities on their own, but they still need to work with enterprise data and custom data sets. It's just much cheaper to develop those applications against them.

One question for those who are deeply involved in this space is, is the base model very generalized and can it do everything? Why do we need custom models and enterprise data?

Frank:

So we have very generalized models that can do poetry, deal with The Great Gatsby's do summaries, do math problems.

But in business, we don't need that, we need a Copilot to get extraordinary insights on a very narrow, but very complex data set.

We need to understand business models and business dynamics. This doesn't need to be computationally expensive, because a model doesn't need to be trained on a million things, it only needs to know very few, but very deep, topics.

As an example. I'm on the board of Instacart, and one of our big customers, like DoorDash and all the other businesses often face the problem that they keep increasing their marketing spend, and a customer comes in, the customer places an order, and the customer either doesn't come back or comes back 90 days later, which is very volatile. They call that churning customers.

This is the analysis of a complex problem because there can be many reasons why customers don't come back. People want to find the answers to these problems, and it's in the data, not in the general Internet, and it can be found through artificial intelligence. This is an example of where there could be tremendous value.

Moderator:

How should these models interact with enterprise data?

Jen-Hsun Huang:

Our strategies and products are various sizes, state-of-the-art pre-trained models, and sometimes you need to create a very large pre-trained model so that it can generate prompts to teach smaller models.

And the smaller model can run on almost any device, perhaps with very low latency. However it does not have a high generalization capability, and the ZERO SHOT (zero sample learning) capability may be even more limited.

So you might have several different types of models of different sizes, but in each case you have to do supervised fine-tuning, you have to do RLHF (reinforcement learning with human feedback) so that it's consistent with your goals and principles, and you need to augment it with vector databases and things like that, so all of that comes together on one platform. We have the skills, the knowledge and the basic platform to help them create their own AI and then connect it to the data in Snowflake.

Now, the goal of every enterprise customer shouldn't be to think about how do I build a large language model; their goal should be, how do I build an AI application to solve a specific problem? That application might take 17 questions to do prompt and eventually come up with the right answer. And then you might say, I want to write a program, it might be a SQL program, it might be a Python program, so that I can do this automatically in the future.

You still have to guide this AI so that he can eventually give you the right answer. But after that, you can create an application that can run 24/7 as an agent (Agent) that looks for relevant situations and reports back to you ahead of time. So our job is to help our customers build these AI applications that are security guarded, specific, and customized.

Ultimately, we're all going to be smart makers in the future, employing employees, of course, but we're going to create a whole bunch of agents that can be created with something like Lang Chain, connecting models, knowledge bases, other APIs, deploying them in the cloud, and connecting them to all of the Snowflake data.

You can operate these AIs at scale and keep refining them. so each of us will be making AI, running AI factories. We will put the infrastructure in Snowflake's database where customers can use their data, train and develop their models, operate their AI, so Snowflake will be your data repository and bank.

With your own data goldmine, everyone will be running AI factories on Snowflake. That's the goal.

03 Although "nukes" are expensive, using the models directly is equivalent to a "10% discount"

Huang Renxun:

We have built five AI factories at NVIDIA, four of which are the top 500 supercomputers in the world, and another one is coming online. We use these supercomputers to do pre-trained models. So when you use our Nemo AI Foundation Service in Snowflake, you're going to get a state-of-the-art pre-trained model with tens of millions of dollars already invested in it, not to mention the R&D investment. So it's pre-trained.

And then there's a whole bunch of other models around it that are used for fine-tuning, RLHF. all of those models are much more expensive to train.

So now you've got the pre-trained model adapted to your function, adapted to your guardrails, optimized for the type of skill or function you want it to have, augmented with your data. So this would be a much more cost effective approach.

More importantly, in a matter of days, not months. You can develop AI applications in Snowflake that connect with your data.

You should be able to build AI applications quickly in the future.

Because we're seeing it happen in real time right now. There are already apps that let you chat with your data, like ChatPDF.

Moderator:

Yes, in the software 3.0 era, 95% of the training costs are already being covered by someone else.

Jen-Hsun Huang:

(laughs) Yes, at a 95 percent discount, I can't imagine a better deal.

Moderator:

That's the real motivation, as an investor, I see very young companies in analytics, automation, legal, and so on, whose applications have realized real business value in six months or less. Part of the reason for that is they're starting with these pre-trained models, and that's a huge opportunity for companies.

Jen-Hsun Huang:

Every company will have hundreds, if not 1,000, AI applications that are just connected to your company's various data. So all of us have to get good at building these things.

04 It was data looking for business, now it's business looking for data

Moderator:

One of the questions I keep hearing from big business participants is, do we have to go invest in AI and do we need a new stack (Stack)? How should we think about connecting to our existing data stacks?

Frank:

I think it's evolving. Models are getting cleaner, safer, and better managed. So we don't really have a clear view that this is the reference architecture that everybody is going to use? Some people will have some central service setup. Microsoft has a version of AI in Azure, and a lot of their customers are interacting with Azure.

But we're not sure what models will dominate, and we think the market will sort itself on things like ease of use, cost. This is just the beginning, not the end state.

The security sector will also be involved, and issues about copyright will be revolutionized. Right now we're fascinated by technology, and the reality of the problem will be dealt with at the same time.

Jen-Hsun Huang:

We are now experiencing the first fundamental change in computing platforms in 60 years. If you just read the IBM System 360 press release, you'll hear about central processing units, IO subsystems, DMA controllers, virtual memory, multitasking, scalable computing forward and backward compatible, and all of these concepts, in reality, are from 1964, and these concepts have helped us scale CPUs over the last six decades.

Such scaling has been going on for 60 years now, but it's come to an end. Now we all understand that we can't scale CPUs anymore, and all of a sudden, software changes. The way software is written, the way software operates, and what software can do is very different from what it used to be. We call the previous software Software 2.0. now it's Software 3.0.

The truth is that computing has fundamentally changed. We see two fundamental dynamics happening at the same time, and that's why things are shaking up dramatically right now.

On the one hand, you can't keep buying CPUs. if you buy a bunch more CPUs next year, your compute throughput won't increase. Because the end of CPU scaling has come. You're going to spend a bunch more money, and you're not going to get any more throughput. So the answer is you have to go to acceleration (Nvidia Accelerated Computing Platform). The Turing Award winner talked about acceleration, Nvidia pioneered acceleration, and accelerated computing is now here.

The other side of that is that the whole operating system of the computer has changed profoundly. We have a layer called NVIDIA AI Enterprise, and the data processing, the training, the inference deployment of that, the whole of that is now integrated or being integrated into Snowflake, so the whole computational engine behind it, from the beginning data processing, all the way to the final deployment of the big model, is accelerated. We're going to empower Snowflake where you'll be able to do more, and you'll be able to do more with fewer resources.

If you go to any of the clouds, you'll see NVI

The power of generative AI to create content, with algorithms to generate images, text, music, and even code and video, opens up endless possibilities for a wide range of industries and has the potential to drive innovation in many fields.

UBS analyst Michael Briest, citing a report by Gartner, the world's leading IT research and advisory firm, said that in the coming years, generative AI will have an impact on "pharmaceuticals, manufacturing, media, architecture, interior design, engineering, automotive, aerospace, defense, medical, electronics and energy industries, medical, electronics and energy industries. Briest said Gartner's report states that "generative AI will impact the marketing, design, corporate communications, training and software engineering sectors by enhancing cross-organizational support processes."

He also said Gartner made five bold predictions about how AI will accelerate enterprise innovation between now and 2027. These predictions are as follows.

1. that by 2025, more than 30 percent of new drugs and materials will be systematically discovered through generative AI technologies;

2. by 2025, the use of synthetic data will reduce the amount of real data required for machine learning by 70%.

3. 30% of external marketing messages for large organizations will be generated by AI by 2025, compared to less than 2% in 2022.

The UBS analysts also noted that it noted that Salesforce recently released Einstein PT, which can send personalized emails to customers on behalf of salespeople, generate specific query responses on behalf of customer service professionals, and deliver targeted content to marketers.

4, By 2027, nearly 15% of new applications will be automatically generated by AI without human involvement, compared to 0% today.

1. By 2026, more than 100 million people will be working alongside their robot colleagues to make their respective contributions to the enterprise.

Will have a significant impact on the labor market

At the same time, this accelerated innovation could also result in significant job losses.

In UBS's view, as with early innovations in areas like printing, electricity or railroads, generative AI could have a significant impact on the labor market.

A recent study published by Goldman Sachs in March this year also shows that the latest breakthroughs in generative AI systems, such as ChatGPT, are expected to bring major disruptions to the global labor market, with 300 million jobs expected to be replaced by generative AI globally, with lawyers and administrators being the most likely to be laid off.

Last month, Ben Goertzel, a cognitive scientist and founder and CEO of SingularityNET, made an even more startling observation. The AI expert said that AI could replace 80 percent of human jobs in the next few years.

① As the CEO of Alphabet, Xander Pichai privately complained that he could not manage some executives. The reason for this may be that founder Page is still in charge.

② The Alphabet CEO is very different from the Microsoft CEO, who has more power and can keep management in line.

③ Google Play earned $10 billion to $12 billion in operating profit in 2022, accounting for 13 to 16 percent of Alphabet's total operating profit.

④ At this year's Google I/O, Google's annual software developers conference, Pichai will likely announce new artificial intelligence features for search, Google Docs and Google Cloud products.

In the field of artificial intelligence, Google has always been like the "West Point" existence, even the OpenAI founding team of R & D staff are basically from Google. But while ChatGPT has exploded out of the gate with its outstanding performance, bringing OpenAI into the public eye and making Microsoft smile with its heavy investment, Google has become the loser in the AI race. Rampant internal bureaucracy, risk averse management, and a draped co-founder are the main causes of Google parent company Alphabet CEO Xander Pichai's inability to embed AI R&D results into existing products such as search. The following is the full text of the article:

Since Pichai became chief executive of Google parent Alphabet in 2019, he has been candid with close friends about the difficulties he faces running the massive Alphabet: a long-standing internal power struggle, strict oversight by regulators, and constant pressure on management from unruly employees. Pichai said in an internal meeting a few years ago that being Alphabet's chief executive had taken its toll on him and that he would hand over the reins in a few years, according to people familiar with the matter.

Pichai shared those feelings with close friends before the panic that has plagued Google in recent months. To this day, the Alphabet marquee is still in Pichai's hands. The emergence of panic has forced him to become more engaged. As advertisers began to cut spending last fall, Google's business began to stagnate. In addition, rivals Microsoft and OpenAI launched a new generation of artificial intelligence services that could threaten Google's dominant position in search and the more than $150 billion a year in advertising revenue associated with it. Suddenly, Google must go after rival Microsoft in areas it once dominated.

"The current situation is calling for the first time for Sandra to become a wartime CEO, and we haven't seen if he's ready for the challenge," said Adrian Aoun. Aoun has sold startups he founded to Google in the past and assisted Google co-founder Larry Page in 2015, restructuring Google into Alphabet.

The problems exposed by this crisis began to worsen during Pichai's tenure at the helm, but they didn't seem to matter until Alphabet's advertising revenue and profits began to shrink: he prefers incremental improvements to products over drastic adjustments; and can tolerate ballooning employee numbers, a lazy corporate culture and an inefficient organizational structure. Perhaps most importantly, despite being CEO of Alphabet, Pichai doesn't have much power over some executives.

For years, Pichai has complained to close friends around him that he can't manage some executives. Pichai has said he can't get Demis Hassabis, CEO of DeepMind, Alphabet's London-based artificial intelligence unit, to prioritize projects or share software code with Brain, another large artificial intelligence unit within Google that often develops the same type of machine learning software as DeepMind. Brain is another large artificial intelligence unit within Google, often developing the same type of machine learning software as DeepMind.

Longtime executives at Google attribute this status quo to a corporate culture similar to that of an academic or government agency. In such a culture, veteran employees may or may not obey orders from their superiors. In addition, several Alphabet employees say Pichai abhors conflict with colleagues.

There may be another reason Pichai can't directly lead the management of Alphabet, which has a market capitalization of $1.3 trillion: Pichai's predecessor, Larry Page, is still draped in power. As co-founder of Google, which developed the money-maker search engine, Page is still on Alphabet's board and controls the company through special stock with another co-founder, Sergey Brin. Although Page has largely stayed out of Alphabet's internal affairs so far, Pichai spoke with Page, who was his boss for a long time, about DeepMind chief executive Hassabis when he wanted his business to be independent of Google, people familiar with the matter said.

01 Very different from Microsoft CEO

Even the recluse Page recognizes the external threat to Google's AI supremacy. Page has long seen Google as a vehicle for the eventual development of general artificial intelligence - computers that can learn on their own and reason beyond humans. In recent months, he and Brin have participated in a number of AI strategy sessions held within Alphabet, a move that reflects Alphabet's extraordinary concern about the threat OpenAI could pose.

Last week, Alphabet announced that DeepMind would merge with Brain to form a new division, Google DeepMind, which Alphabet said would "dramatically accelerate our progress in artificial intelligence." While it may seem logical to integrate the teams, which have been developing similar software in parallel, the move came as a shock to many Google employees because the two teams have different cultures, and because for years Brain's head Jeff Dean, who has been in competition with DeepMind's Hassabis, chose to to step aside.

Previous media reports have claimed that Page has attended several internal company meetings. According to an employee, shortly after the recent Alphabet board meeting, Google announced the decision to merge its AI teams, with Hassabis to head the new combined team. This timing suggests that the final decision may come from the board level, rather than Pichai.

Pichai's power at Alphabet contrasts with Microsoft CEO Satya Nadella's power at Microsoft. Nadella seems to be able to keep his management team in lockstep as he ties Microsoft's future to that of OpenAI through the partnerships he has built by investing tens of billions of dollars. This means that some Microsoft executives and managers have had to make sacrifices: Microsoft is allocating computing power company-wide for training AI models so that efforts related to OpenAI, including embedding OpenAI's technology into Microsoft's core products, can have a better chance of success. Microsoft's Azure cloud business is funding OpenAI's snowballing computing infrastructure and AI development. After Nadella lost patience with the sheer number of AI researchers within Microsoft, he outsourced AI development to OpenAI.

02 Standing the last shift

In the years since he became CEO, Pichai has continually told some colleagues that he misses the way things used to work. But one person who works closely with him says Pichai plans to lead Alphabet through its recent financial and competitive challenges. Pichai's publicists have thrust him into the spotlight to showcase Google's work developing the next generation of artificial intelligence products. He recently appeared on CBS's "60 Minutes" and a New York Times podcast.

Whether Pichai can stay on may not be up to him. According to current and former Alphabet employees, the general consensus across the company is that he may not be the right person to continue as CEO at this time. Given Alphabet's relatively strong stock performance since he took over, investors don't seem to be calling for a change of hands at Alphabet, but there are still some activist investors clamoring for him to cut costs. According to Koyfin, the market has sent warning signals in recent weeks because Alphabet's stock valuation (dynamic price-to-sales ratio) is lower than Facebook parent Meta for the first time in a year.

As Pichai has repositioned Alphabet on artificial intelligence, some executives have been promoted. Some media reports say that the promoted executives include Thomas Kurian, head of Google Cloud. Before the promotion, Kurian just took over Google's team responsible for designing a special chip to support artificial intelligence. Earlier, Kurian called for more firepower to fight Microsoft in the business of selling artificial intelligence services to cloud customers. With Google's advertising revenue growth stagnating, the cloud business is more important than ever. In its latest quarterly earnings report, Alphabet disclosed for the first time that its cloud business has achieved an operating profit.

In addition to Kurian, the executives promoted include Hassabis and James Manyika, a former Obama-era administration official and consultant at consulting firm McKinsey, and one of several executives who appeared on "60 Minutes" with Pichai. Manyika just took over a roughly 2,000-person research group from Brain's original head, Dean, and is deeply involved in Alphabet's artificial intelligence product strategy.

03 The most disciplined executive

A former software engineer and business consultant, Pichai joined Google as a general product manager in 2004, but over the next decade he became Page's designated successor. Pichai rose through the ranks because he was the company's lowest-profile, most compliant executive during a period when Page was purging executives who he thought were too divisive or morally corrupt.

When Page reorganized Google into Alphabet in 2015 and became CEO of the parent company, Pichai was promoted to CEO of the subsidiary Google. As the parent company, Alphabet oversees Google and various subsidiaries, setting goals for long-term projects such as driverless cars and smart cities. By separating from Google, these technologies are expected to be better developed.

In the years since Alphabet was founded, Google has relied on its search engine and Android mobile operating system to allow steady growth in revenue and net profit. Starting in 2015, Google's advertising business has benefited because of machine learning techniques developed by Dean's team and other researchers hired by Google from academia.

In this article, we will explore the beauty and importance of the rainforest, the threats it faces, and the steps we can take to protect it. We will also discuss how you can get involved in our community and receive a free NFT by subscribing to our account.

The Beauty and Importance of the Rainforest

The rainforest is one of the most beautiful and important ecosystems on the planet. It is home to a vast array of plant and animal species, many of which are found nowhere else in the world. The rainforest also plays a vital role in regulating the Earth's climate by absorbing carbon dioxide and producing oxygen.

However, despite its importance, the rainforest is under threat from a variety of sources. Deforestation, climate change, and pollution are just a few of the many threats facing this vital ecosystem.

Protecting the Rainforest: A Call to Action

As environmental defenders, we are committed to protecting the rainforest and its delicate ecosystem. Some of the steps we can take to make a difference include:

1. Support Sustainable Agriculture:

By supporting sustainable agriculture practices, we can help to reduce the impact of deforestation on the rainforest. This can include buying products that are certified as sustainably sourced, and supporting farmers who use environmentally friendly practices.

2. Reduce Your Carbon Footprint:

Reducing your carbon footprint can help to reduce the impact of climate change on the rainforest. This can include reducing your use of fossil fuels, eating a plant-based diet, and using public transportation or carpooling instead of driving.

3. Support Conservation Efforts:

Supporting conservation efforts can help to protect the rainforest and its delicate ecosystem. This can include donating to organizations that work to protect the rainforest, volunteering your time to help with conservation efforts, or advocating for policies that support conservation efforts.

Join Our Community: How to Get Involved

If you're interested in learning more about the importance of protecting the rainforest and the steps you can take to make a difference, we invite you to join our community. Our account is dedicated to sharing our knowledge and insights on the importance of protecting the rainforest and the steps we can all take to make a difference.

We are also giving away free NFTs to our subscribers. By subscribing to our account, you can mint your own free NFT, which will give you exclusive access to our community and our insights into protecting the rainforest.

https://opensea.io/assets/ethereum/0x3Ea523578F54f6e9f924b757962f6D7c95319B7d/0

In conclusion, the rainforest is one of the most beautiful and important ecosystems on the planet, but it is under threat from a variety of sources. As environmental defenders, we are dedicated to protecting the rainforest and its delicate ecosystem for future generations. By taking steps to support sustainable agriculture, reduce your carbon footprint, and support conservation efforts, you can help to protect the rainforest and make a difference. We invite you to join our community and share your insights and analysis with us. By subscribing to our account, you can also mint your own free NFT and gain exclusive access to our community and insights.

Industry Trend

On March 17, in order to promote the innovative development of Beijing's Internet 3.0 industry, with the consent of the municipal government, the Beijing Municipal Science and Technology Commission and the Beijing Municipal Bureau of Economy and Information Technology of Zhongguancun Science and Technology Park Management Committee issued the notice of "Work Plan on Promoting the Innovative Development of Beijing's Internet 3.0 Industry (2023-2025)". Internet 3.0, as a new form of future Internet industry development, is accelerating under the impetus of virtual reality, artificial intelligence, blockchain and other information technologies, and will have a transformative impact on future technological and industrial development as well as social and economic patterns. In order to seize the opportunity of the new round of scientific and technological innovation and industrial change, and promote the innovative development of Beijing Internet 3.0 industry, this work plan is formulated. (Beijing Municipal Science and Technology Commission, Zhongguancun Science and Technology Park Management Committee)

Wuxi released a three-year action plan for metaverse innovation and development to strengthen metaverse core technology research

On March 16, the "Wuxi City Three-Year Action Plan for Metaverse Innovation Development (2023-2025)" was released, aiming to make Wuxi City a highland of metaverse innovation, meta-industry development and metaverse industry by 2025. By 2025, Wuxi will become a highland of "meta-technology" innovation, "meta-industry" development, and "meta-work" creation, with its technological innovation capability, core industry scale and application demonstration at the leading level in China, and the meta-universe becoming a new engine to boost the development of Wuxi's digital economy. The Action Plan specifies key tasks, including strengthening theoretical and technological innovation breakthroughs in metaverse, promoting the development of metaverse eco-industry clusters, creating demonstration projects of metaverse application scenarios, and optimizing the ecological environment for metaverse innovation and development. (Wuxi City Bureau of Industry and Information Technology official website)

White Paper on "China's Network Rule of Law Construction in the New Era": actively promoting the in-depth application of modern technologies such as blockchain in litigation services and other fields

On March 16, the State Council Information Office released a white paper on "China's Network Rule of Law in the New Era," pointing out that China is actively exploring new paths, new areas and new models for the in-depth integration of judicial activities and network technology, so as to "speed up" social justice.

The white paper points out that China is actively promoting the in-depth application of modern technologies such as big data, cloud computing, artificial intelligence and blockchain in areas such as litigation services, trial execution and judicial management, and building a cyber justice model with Chinese characteristics on an early and pilot basis. Encourage local courts to explore new Internet trial mechanisms with regional characteristics, taking into account the development of the local Internet industry and the characteristics of Internet disputes. Vigorously promote the work of digital prosecution, adhere to the big data to empower legal supervision, systematically integrate all kinds of case data, and actively explore the construction of big data legal supervision model and platform, and strive to promote the combination of case-by-case supervision and case management supervision, to improve the quality and efficiency of legal supervision in the new era. (Xinhua News Agency)

General Administration of Market Supervision: 2022 digital collections and other new models of problems rise, supervision more difficult

Recently, the State Administration of Market Supervision said that some new situations of infringement and new problems of rights protection are gradually exposed, reflecting the new pain points of consumer rights protection. 2022, digital collections (NFT) and other new models of problems are on the rise and supervision is more difficult, with 59,700 related claims (only 198 cases in the previous year), mainly focused on non-delivery, non-refund, malicious price inflation, charging high fees, arbitrarily blocking consumer accounts, etc. (Xinhua Finance)

Virtual Reality and Metaverse Industry Alliance's "Technical Requirements for Trusted Virtual Human Generation Content Management System" Standard Development Work Launched

Recently, in the face of the accelerating market trend and the accompanying risks and challenges, guiding the development of the technology industry with credible principles has become a necessary path to promote the safe and controllable, sustainable and high-quality development of virtual humans. Relying on the Virtual Reality and Metaverse Industry Alliance (XRMA), China Academy of Information and Communication Research (CAICR), led by Shangtang Technology, and edited by OPPO, Baidu, Northern Polytechnic University, Erzhusan, VIVO, Virtual Dynamic Point, You Chain Era and Soul APP as the first participating units, jointly initiated the standard development work of "Technical Requirements for Trusted Virtual Human Generation Content Management System", which has passed XRMA The standard has passed the project review within XRMA. Now, we are going to launch the plan of soliciting participation units and experts for the development of the standard, and the units and experts who make outstanding contributions to the preparation and implementation of the standard will also be included in the standard as core preparation units and experts. (China ICT public number)

Guotai Junan: Multimodality is the inevitable trend of GPT series development and the basis for diversified applications to be implemented

Recently, Guotai Junan pointed out that multimodality is an inevitable trend in the development of GPT series and also the basis for the landing of diversified applications. Considering that a significant portion of today's information data is presented in the form of images and videos, GPT-4 with image and video processing capabilities will obtain information from more complete sources and present it in the form of multimedia, effectively improving user experience. In the long term, multimodality will open up the visual direction, image generation and video creation capabilities, which will assist GPT-4 to achieve further broadening in various business models, thus realizing multimedia interaction. (Guotai Junan)

Targeting 1 million new "metaverse population", Jiangxi launched the "One Tree Landscape" metaverse "citizenship" project

Recently, in order to implement the national policy requirements and seize the opportunity of metaverse development, Shangyu County, Ganzhou City, Jiangxi Province, based on local resource endowment, has explored a way to develop the deep integration of digital economy and real economy. Combined with the advantages of local natural resources, the "One Tree Landscape" metaverse "citizenship" project is about to issue metaverse "ID cards" for people all over the world, with the goal of adding 1 million people to the metaverse. The goal is to add 1 million people (metaverse "citizens"). Starting from March 12 (Arbor Day), users who purchase the digital collection "One Tree Landscape" through the platform channel or offline activities will receive a metaverse "citizenship" ID card, and after having this identity, they will enjoy the future "citizenship" card one after another. The "one tree scenery" platform of various online consumption and free rights and interests, at the same time will be open one after another on the whole area of cultural tourism services in Shangyu County and a number of free and discount rights and interests. (Phoenix Shenzhen Technology)

Investment and Financing

Bigpen VR Completes New Round of Strategic Financing

Recently, Enterprise Search shows that VR manufacturer BigPen VR has completed a new round of financing for an undisclosed amount from Qingdao Wenxin Capital Operation Co.

Up to now, DaPeng VR has completed 11 rounds of financing, with investment institutions including Kaixin Network, Xunlei Network, Aofei Entertainment and Huaqiang Capital. According to IDC data, the annual VR/AR global shipments in 2022 were 8.8 million units, and DaBen VR ranked third in the world, after Meta and PICO. 2022, DaBen VR released a 6DoF PC VR product, DaBen E4, to further lay out the C-terminal market.

AR HUD Technology Provider Envisics Raises Over $50 Million in Series C Funding

Recently, Envisics, a provider of AR HUD technology, announced that it has received more than $50 million in Series C strategic financing, led by Mobis, with additional investment from new strategic shareholders InMotion Ventures and Stellantis.

Envisics' holographic technology is understood to be used to create augmented reality flat-screen displays (AR HUDs), a display technology considered a key feature of modern cars that can also provide drivers with better situational awareness, thereby improving road safety. Mitchell Caplan, chairman of Envisics, said the new financing will provide Envisics with the runway it needs to realize the commercial potential of dynamic holographic technology in the automotive industry.

Big Brother Remarks

Paul Chan: Hong Kong to introduce licensing system for virtual asset service providers in June

Recently, the Financial Secretary of the Hong Kong SAR Government, Mr. Paul Chan, said that the virtual asset industry has been developing steadily in Hong Kong over the past few years. Under the principle of "same activity, same risk, same regulation", Hong Kong will introduce a licensing system for virtual asset service providers in June this year. Mr. Chan added, "We believe that by establishing a suitable regulatory licensing regime to exclude the black sheep in the industry, it will be conducive to enhancing the overall image of the industry and promoting the healthy, orderly and responsible development of the virtual asset industry."

Zhou Hongyi: GPT-4 will set off a new industrial revolution, domestic catch-up should not give up core technology research and development

Recently, 360 founder Zhou Hongyi said, I think GPT-4 will set off a new industrial revolution, bringing new social division of labor and creating new application scenarios. The intelligent level of human society will be improved comprehensively. By then, all digital apps, software and websites are worth redoing with ChatGPT.

Enterprise News

China Mobile and ZTE Complete the Industry's First Digital Twin Application for XR Services in Wireless Networks

China Mobile Research Institute (CMRI) and ZTE have recently completed the industry's first digital twin modeling of wireless networks for XR services in Jiujiang, Jiangxi Province. In the future, China Mobile and ZTE will continue to cooperate to carry out new service assurance and network strategy optimization based on the wireless network digital twin platform, and promote the development of intelligent applications in wireless networks. (Speedway Metaverse Research Institute)

PICO will launch PICO 4 OS 5.5.0 system public test campaign.

To that end, Nansen has compiled a list of safer stablecoin Yield Farming strategies to help you minimize risk while maintaining passive income during relentless bear markets.

Stablecoin and Yield Farming First, stablecoins are not inherently risk-free assets. Investors first need to understand the general and asset-specific risks associated with each type of stablecoin and avoid treating these digital assets as "real" dollars. Despite the various risks associated with stablecoins, they remain attractive due to their ability to be used in Yield Farming strategies.

Since Yield Farming means putting assets into DeFi applications to collect returns, stablecoin Yield Farming intuitively means generating passive income on stable crypto assets through two types of lending activities: money market lending and lending as liquidity to decentralized exchanges.

Yield Farming on DEX Among the many DEX designs, stablecoin exchanges such as Curve offer minimal asset requirements and low transaction fees for trading similar assets (e.g., between two stablecoins such as USDC and DAI), which require highly liquid pools of stablecoins.

These platforms deploy a variety of incentives focused on stable assets, which often translate into opportunities for passive income seekers. Such exchanges include

Curve Finance

The technical or conceptual progenitor of the stablecoin exchange, Curve is the most used and liquid stablecoin trading service platform in the market today. The platform is seen as one of the most reliable sources for putting stable assets to work. Some have even referred to the Curve stablecoin liquidity pool as a savings account for cryptocurrencies.

Providing assets to Curve allows depositors to receive a portion of the transaction fees generated in the pool, as well as additional CRV token emissions. The most traded of Curve's stablecoin pools is 3pool, consisting of DAI, USDC, and USDT (three of the four most liquid stablecoins on the market). At the time of writing, 3pool offers 0.10% of the stablecoin APY, which varies daily with trading volume, and 0.2% of the token APR (CRV Reward), which depends on the reward rate, price, and revenue earned through pledging.

While these returns are low, Curve has other more rewarding stablecoin pools, and 3pool is widely considered to be safer because the pool and the stablecoins traded on it are strongly "stress tested". Curve pools on other EVM chains and Rollup can be accessed, but most liquidity is concentrated on the main ethereum network.

Ellipsis Finance

A fork of Curve, Ellipsis Finance has the same technical logic as Curve and offers similar services on the BNB Chain, but with a different range of stablecoins in focus (e.g. BUSD and USDD). Like Curve, investors can deposit stablecoins as liquidity and earn interest from trading activity. Because Ellipsis is on the Coin On chain, the transaction fees for accessing assets and collecting rewards are significantly lower than Curve Finance.

There is a difference between the capital efficiency of Ellipsis and Curve due to different network effects. Ellipsis Finance's most liquid pool ($32 million in TVL and $215,000 in volume) offers a 0.06% APR benchmark compared to Curve's 3pool (over $950 million in TVL and $78 million in volume), which is a 0.10% APR benchmark. Ellipsis' low usage is compensated by higher transaction fees and a high EPX token reward - 1.47% at the time of this writing.

Farmer needs to remember that the return on stablecoin trading is variable and varies with the day's trading volume. Therefore, calmer markets and lower project token prices can translate into lower dollar designated returns.

Alternatives to the platforms mentioned above include several other stablecoin exchanges, such as Platypus Finance on Avalanche, which focuses on major stablecoins, and Saber on Solana, whose most liquid pools include USDC and UXD.

Other DEX Spot DEXs typically do not provide the same incentive for users to buy fixed assets as stablecoin exchanges. However, investors can also lend stablecoins to DEXs such as Uniswap, Sushiswap, Pancakeswap, TraderJOE, Quickswap, Serum, and Osmosis. since these applications do not provide unilateral liquidity supply, users need to deposit two assets to become a liquid market maker (e.g. DAI & USDC, USDT & BUSD).

A third alternative to stablecoin and multifunctional DEX is cross-chain bridges like Synapse Protocol and Hop Protocol, which also allow for the provision of a single asset as liquidity for stablecoin-like transactions.

Check out Nansen's liquidity mining dashboard to find the best opportunities in the liquidity supply market.

Crypto Lending The second most popular way to earn passive income on stablecoins is to use decentralized cryptocurrency marketplaces like Aave and Compound. Just as stablecoin exchanges often reward lenders with stablecoin and native Dapp tokens, cryptomarkets typically compensate depositors with generated revenue and governance tokens.

For example, when a user deposits their USDC into Aave's loan pool, they receive a corresponding token, aUSDC, a 1:1 liquid synthetic asset redeemable for USDC. As the loan position expires over time, the fees collected from the borrower are allocated to the user's wallet on a pro-rata basis, resulting in a steady increase in the aUSDC balance, which can be redeemed at any time for the underlying stablecoin. The APY on USDC and USDT on Aave is currently 0.69% and 1.96%, respectively, and varies with borrowing rates and utilization. In addition, users can earn AAVE rewards by depositing into specific pools.

In addition to these two blue-chip lending programs, investors can also lend their stablecoins to cryptocurrency markets that fall under different licenses.

Rari's unlicensed Fuse pool allows users to create custom money markets using the assets they want and lending parameters such as pledge factors and interest rate models determined by the pool creator. This allows both the most popular stablecoins and the long-tail coins with relatively high interest rates to be profitable. However, this comes with the endemic risk of instability.

Decentralized Credit Protocol Goldfinch stands alone in the licensing space with its risk-weighted asset (RWA) based secured crypto lending service. Anyone can offer their USDC to earn interest, but only vetted parties with the proper credit limit can borrow from the loan pool. To date, the Premium Pool offers lenders 7.81% of their USDC APY and another 9.42% in return for GFI tokens.

On top of these, there are a number of cryptocurrency markets and DEXs built on top of, or working in coordination with, less liquid niche stablecoins. Examples include USDJ and the JustLend lending market, OUSD and Origin Protocol, agEUR and Angle Protocol, and most recently GHO on Aave.

However, these niche platforms should insist on better risk management, as the applications and stablecoins they serve are not as "stress-tested" and scrutinized as their blue-chip counterparts. As the most extreme examples of these platforms (such as Anchor Protocol and UST) have crashed with significant losses to investors, we recommend caution for those looking to convert their stable assets and use them on these platforms.

Conclusion Yield Farming is a double game. The way you operate determines whether it is an active or a passive income generating instrument. Investors can look only at returns, constantly researching better yields and allocating their money between different agreements, or go heavy on individual agreements. The former is riskier and requires effort, but may yield higher returns, while the latter brings more security and inner peace of mind. Try to find the risk-appetite strategy that works best for you. Once you've made up your mind, using the Nansen tool can help you find the best market opportunities for returns.

Noisy Trading Lacking Upward Momentum for Now Risky assets climbed after Fed Chairman Powell said he would not add 75 basis points, with trading volume and real volatility spiking around the announcement, but trading volume quickly fell back within 24 hours.

The intraday "relief rally" in cryptocurrencies and stocks on Wednesday was "trader noise," according to Mike McGlone, a commodity strategist at Bloomberg Intelligence. (Noisy traders are usually non-professionals who act illogically and trade with incomplete or inaccurate data.)

At the time, the Glassnode team also cautioned that bitcoin prices remain range-bound and continue to lack any clear macro momentum in either direction, that correlations between bitcoin and traditional markets remain near all-time highs, and that the broader perception of bitcoin as a risk asset remains a significant headwind.

Intertwining with Global Economic Factors Adding to Uncertainty Following the Fed's rate hike meeting, on May 6, Robert Holzmann, a member of the European Central Bank's Management Board, said that the central bank will discuss a rate hike at its June meeting and may decide to raise rates once.

GSR Institutional Crypto Trader noted that bitcoin has had its ups and downs in correlation with stocks, especially during major macro events such as the Federal Open Market Committee meeting. In terms of sentiment, it is more important to see how the cryptocurrency market performs at the close of the equity market.

In addition, there is no relaxation in regulation of the crypto market, as recently, European regulator MONEYVAL listed cryptocurrencies as one of the anti-money laundering threats. In addition, the Central Bank of Argentina (BCRA) announced that banks in the country are banned from offering cryptocurrency services to their customers. The BCRA's statement said that banks are prohibited from providing services for any digital assets that are not regulated by the central bank, a move that amounts to a de facto ban as digital assets are currently not regulated by the Argentine government.

Market analysis suggests that several factors, including rising inflation, geopolitical crises, crypto regulation and shifts in U.S. monetary policy, continue to drive additional short-term volatility in cryptocurrency and equity markets. In recent months, the crypto market has increasingly tracked the stock market, which has made it more intertwined with global economic factors.

On May 6, the three major U.S. stock indices closed sharply lower, with the Nasdaq down 4.99%, the S&P 500 down 3.55%, and the Dow down 3.11% And clearly, bitcoin prices have been affected by the correlation.

Institutional Investors' Mindset Dominated by Uncertainty The current crypto market sentiment is not positive. According to Coinbase analytics platform Skew, implied volatility (a measure of investors' willingness to buy BTC options) has fallen to its lowest level (3.1%) since the beginning of 2019. The metric measures how much options traders expect to pay in the near future.

Michael Saffai, a partner at Dexterity Capital, a crypto asset trading firm, said near-term uncertainty continues to dominate the mindset of institutional crypto investors. The recent liquidation could exacerbate the pullback, but the asset still has a solid $30,000 floor, so I don't think we'll see a massive pullback similar to what we saw in 2020 and 2021.

Also, according to the latest Web3 report from blockchain analytics firm Chainalysis, NFT saw explosive growth in 2021, but the market leveled off heading into 2022, maintaining a growth attitude in January but falling into a slump in February before starting to recover in mid-April. As of May 1, more than $37 billion has been injected into the NFT market. The number of active NFT pools on OpenSea has also continued to grow since March 2021 and now exceeds 4,000. However, the growth of institutional investors has not continued. NFT purchases by institutional investors grew weekly between late November 2021 and mid-February 2022, but then abruptly declined and institutional NFT activity has not yet reached the levels seen in winter 2021.

In addition, according to the latest data from Coinshares, outflows from the market in one month due to institutional investor withdrawals totaled more than $339 million. While the market saw similar volatility earlier this year, the study suggests that this volatility did not reverse that trend. According to Coinshares, it started the year at about $467 million, a difference of $128 million. According to the report, bitcoin funds accounted for the majority of recorded withdrawals.