jakiro
Author: Evie Source: X, @0xEvieYang
The market has been terrible these two days. Let’s take a look at the AI transfer.
As the saying goes, “speculating on new things, not old things”, the currency circle has been looking for new narratives.Since last year, many Web3+ AI projects have been released, and more than 50 AI projects have been completed this year.
This year, I have also purchased popular AI concept tokens such as $WLD and $LPT.However, I have always been curious about how mainstream AI is developing?What is the difference between Web3+AI and pure AI?What other opportunities does Web3+AI have?
It is said that “it is difficult for people to make money beyond cognition.”In May, San Francisco held a GenAI conference with 10,000 people participating @genaisummitsf. I took the opportunity to stay in the United States for one month and visited local AI, investors, entrepreneurs, researchers @FinanceYF5, etc.Next, I will share my experiences and thoughts with you.
Due to limited space, this tweet is mainly focused on the development of mainstream AI circles, including:
AI investment and financing situation
AI entrepreneurship atmosphere
AI segmentation track situation
China-US AI development
AI investment and financing situation
I looked at projects with financing of more than US$50 million this year and found that most of the projects are 2B, including vertical categories such as health/medical, transportation/driving, finance and finance, as well as tools to improve organizational efficiency; secondly, cloud platforms are cloud platformsOr computing service provider; there are very few applications on the 2C side.
Regarding this issue, my opinion is that the death rate of C segment is very high at this stage, the number of existing AI users is far from enough to support the cost of C-end applications, and the large amount of high-quality data required for C-end applications is in large manufacturers.In your hands, not start-ups.Based on this, some investors even believe that 90% of the C-end opportunities are in large manufacturers.
AI entrepreneurship atmosphere
My feeling in the Bay Area is that even breathing is AI-smelling.During the roadshow of the GenAI conference, some projects even started pitching with just a simple idea. It can be seen that the entrepreneurial market is still relatively inclusive in this field, which feels like “mass entrepreneurship and innovation” back then.
However, if you really want to get out of many AI projects, it is much more complicated than you imagined, and it still requires more technology, background, and resources.The current celebrity AI projects and teams are all of North American top universities + large factories, or are successful entrepreneurs in succession.
As for the organizational structure of the project, one feature I have observed is “inverted pyramidization” – that is, there are more high-level members, and there are fewer Junior engineers.
Development of AI segmented track
Computing Power: The market now has a large demand for Nvidia, but the supply is very limited, and large companies also need to grab GPUs.The competition among big model companies is a cruel money competition. You need more capital to buy more cards and grab more talents.In addition to the supply and demand issues of GPU, the reduction in energy consumption is also a need to be solved.
Data: Big model development requires powerful GPUs, but at the same time, data is also attracting more and more attention as another key resource, so some top AI labs are also competing for more valuable data, which they will spend.A lot of money is used to buy data, find experts to generate data, or work with companies like Scale AI to mark data.
Some researchers predict that high-quality data will be used up in 2026.Therefore, the importance of synthetic data is increasing.In 2024, 60% of the data used to train AI is expected to be synthetic data.
Model: I have heard different opinions on which is better for open source or closed source models.Some investors are very optimistic about open source and believe that open source can attract contributors to participate. Whether it is a large company or a startup, a lower-cost model may appear under the open source model.Currently, there are models that can reach the level of ChatGPT 4.Another view is that most open source models have not been verified by computing power, and the market does not pay for it, so it is definitely more support for closed source talents and resources.
Based on open source logic and nesting a Web3 business model, everyone can contribute to a set of models and share the model benefits based on the contribution.There are projects doing similar things now, but whether it is feasible, I won’t go into it here for now.
In addition, most relatively mature models are supported by cloud service companies. For example, in the recent round of financing of 1 billion US dollars in the Dark Side of the Moon, Alibaba, as the leading investment, part of which is investing in computing power.
Enterprise software service companies like Salsefore also have their own AI team of hundreds of people, and their AI directly serves their products.
Application: Chatbot is a must-fight place for major manufacturers. There are relatively few major manufacturers in the search field, mainly Microsoft, and NewBing is currently in a monopoly position.
Although Apple’s stock price has fallen after announcing its AI plan at this year’s developer conference, I personally look forward to the combination of Apple and AI. After all, Apple is the most used electronic device in daily life, and it has its ownThe models, chips, clouds, and massive data form an ecosystem. Each link is optimized a little, and it will be very strong when superimposed.
Development of AI in China and the United States
In terms of the development of AI in the United States, innovation is still in the Bay Area.AI startups in Silicon Valley receive much more venture capital than other regions.AI in New York is mainly focused on application, and some companies are using AI to replace or assist in the work of paralegals.
I met a friend who was doing AI consulting services in New York. They were helping some traditional companies with AI system solutions.The integration of AI and corporate workflow is irreversible. I feel that in a few years, juniors in the consulting, auditing, lawyers and other industries will face considerable pressure to lay off employees.
The big models are mainly concentrated in the United States, followed by China and Europe; the number of big models released in the United States last year is 3 to 4 times that of China.The domestic big model – Dark Side of the Moon, announced this year that it has raised 1 billion US dollars, and major manufacturers such as Tencent and Alibaba have also entered the market, which is considered to be the “power of the whole country” to support their big model.
