AI fund year-end review: understand the AI trends in 2024 in one article

Understand the first opportunity in 2024.

Written by: Otter

** Edit: Zuri**

Coatue - The best AI is yet to come

Coatue is a New York-based hedge fund that has attracted industry attention for its persistent bets on the TMT track. It was founded in 1999 by Philippe Laffont (Philippe Laffont) after he left Tiger Fund. Its management scale has spanned from $45 million to $19 billion. Coatue released a research report titled ‘Al: The Coming Revolution’ (“AI: The Coming Revolution”) in November, discussing the revolutionary impact of AI and how AI technology is changing our world.

Coatue & Philippe Laffont |图源：Institutional Investor

**Coatue believes that the best era of AI has yet to come. In the future, the development of AI will be more popular, scalable and practical, and shared the following views and predictions: **

The programming language of the future is undoubtedly English. In the future, AI will make it easier for more people to get started programming, debugging and deploying software.

AI on devices will be widely used. In the future, everyone will have a powerful AI assistant in their pocket, and one billion Apple users will become AI users.

Scaling AI models can be an engineering challenge, and the number of OpenAI engineers has more than tripled since the release of GPT-3.

Multimodal large models are a cutting-edge innovation direction, and there will be new developments in infrastructure. Coatue pointed out that a new architecture “Hyena” proposed by researchers from Stanford University and Mila is attracting attention in the field of natural language processing and is considered to have the potential to subvert the existing attention mechanism system. Although many large models are currently based on Transformer, which requires high computing power, it may be gradually replaced by architectures that require less computing power.

Radical Ventures 10 Predictions Related to Artificial Intelligence (2024 Edition)

Radical Ventures was established in 2018 and focuses on investing in early-stage projects in cutting-edge fields such as AI, robotics and quantum computing. The two founders joined forces with Geoffrey Hinton and others to launch the Vector Institute for Artificial Intelligence in 2017. It is currently a research institution that brings together more than 1,700 of the world’s top artificial intelligence researchers. In addition, Li Feifei is currently serving as the agency’s Scientific Partner.

Radical Ventures partner Rob Toews has always had his own column in Forbes. On December 21, he released Radical’s 10 predictions around AI trends**:**

Radical Ventures & Rob Toews | Source: TED Talks

**NVIDIA will accelerate its entry into cloud computing. **As cloud computing providers are developing their own native AI chips, NVIDIA has also launched a cloud service called DGX Cloud. Radical predicts that Nvidia’s relationship with large cloud computing providers will become more complicated as it enters 2024.

**Stability AI will go out of business. **Stability has been undergoing large-scale layoffs in recent months. In addition, Stability’s monthly expenditure is US$8 million, and the revenue it brings is a drop in the bucket. It is estimated that the investment of US$50 million will not last until the end of 2024. While Stability has begun looking for acquirers, so far not many have shown interest. Therefore, Radical predicts that Stability will completely collapse in 2024.

**The terms “large language model” and “LLM” will no longer be common. **Although the initial generative AI models (such as GPT-3) are pure text models, as AI becomes more and more modal, “large language models” and “LLM” will become more and more The less accurate. Therefore Radical predicts that by 2024, as models become more and more multidimensional, the terms used to describe them will also become more and more multidimensional.

**The most advanced closed models will continue to lead the way. **Although the AI open source community is highly active, most cutting-edge AI model developers (such as OpenAI’s GPT-4) are closed source. Many open source advocates believe that the performance gap between closed and open models is closing, but Radical takes the opposite view that the best closed models will continue to significantly outperform the best open source models in 2024 and beyond.

**Fortune 500 companies will create a new position - chief AI officer. **Radical predicts that next year, large companies will appoint a new role of “chief AI officer” to lead the company’s AI strategy. President Biden’s recent AI executive order also explicitly requires each federal government agency to appoint a chief AI officer.

**Alternatives to the Transformer architecture will appear. **The Transformer architecture proposed by Google in a seminal paper published in 2017 is the mainstream paradigm of today’s AI technology. However, the next generation of AI architecture represented by Mamba is beginning to emerge, which is superior to Transformer in different aspects. Radical predicts that alternatives to Transformer will appear in 2024.

**Cloud providers’ strategic investments in AI startups may be questioned by regulators. **Since this year, Microsoft, Alphabet, and NVIDIA have all invested heavily in some AI startups that they are optimistic about. These investments are motivated, at least in part, by ensuring that these high-growth AI startups can later become long-term customers, which can involve an important gray area in accounting rules. Radical expects that next year the U.S. Securities and Exchange Commission will conduct more stringent scrutiny of round-trip capital inflows in AI investments, which may affect the AI investment and financing environment.

**The relationship between Microsoft and OpenAI will begin to show cracks. **Microsoft and OpenAI have different organizational structures and different visions for the future of AI. As OpenAI aggressively expands its business, it will increasingly compete directly with Microsoft for customers in the future. As far as Microsoft is concerned, for its more diversified development, it will also cooperate with OpenAI competitors in the future.

**Venture capital may turn to the cryptosphere in 2024. **Many well-known venture capitalists, entrepreneurs, and technologists in AI have a soft spot for cryptocurrencies during the 2021-2022 bull market. If crypto asset prices surge back next year, some investors may follow the enthusiasm in this direction, while investment enthusiasm for AI will decline.

**The copyright issue for AI-generated content is still the “Sword of Damocles”. **The fact that currently leading generative AI models are trained on large amounts of copyrighted content could trigger huge legal liabilities in the future and change the industry landscape. Therefore, Radical points out that it would not be surprising if generative AI models like GPT-4 and Midjourney are found by courts to infringe copyright, and the companies that build these models are held accountable to the owners of the intellectual property that trained these models.

a16z Year-End Report: Disruptive Trends in 2024

On December 6, the top VC institution in the United States a16z released a trend forecast report titled “Big Ideas in Tech for 2024”, which covers American society, medical care, consumer technology, cryptocurrency, financial technology, In the fields of games, growth-stage technology, infrastructure and enterprise services, a16z partners are invited to look forward to 2024. The predictions related to AI are as follows:

Photo source: a16z

**AI will facilitate the rapid development of the medical and health field. **Judging from the characteristics of the industry, while other industries have used non-AI software to run efficiently, the medical and health industry still relies on pagers, fax machines and a large amount of manual data entry, so the emergence of AI tools will greatly satisfy The demand for increased productivity in the medical and health field and skipping existing software are revolutionizing medical technology.

**Consumer technology field: voice-first applications, professional AI assistants, and AI will shine in the fields of education and artistic creation. **By 2024, voice applications will become more popular, and the email function of AI voice interaction will be integrated into Gmail; the professional ChatGPT AI assistant will be promoted to various vertical fields, such as one specially built for researchers AI academic assistant, AI writing generation tool for journalists, and AI rendering platform specially designed for designers, etc. By 2024, we may see breakthrough AI educational tools carefully designed for children. At the same time, as AIGC technology reduces the marginal cost of creation to close to zero, new consumer behaviors will also emerge in the field of artistic creation.

**AI and blockchain technology will be deeply integrated. **Currently, AI models like ChatGPT can only be trained and run by a handful of tech giants, but with blockchain technology, it is possible to create multilateral, global, decentralized markets where anyone can contribute— —and get paid for it, which greatly reduces development costs and thus promotes the development of AI technology.

**Fintech sector:**AI technology will improve the efficiency of financial professional services and change the way the banking and trading industries work. LLM technology can be used to capture data that is difficult to process with traditional financial software.

**Game field: AI will create games that never end. **By 2024, we will see the first novel game systems and mechanics developed using large-scale models of AI. AI will revolutionize the nature of gaming—the game world itself will no longer be rendered, but instead generated at runtime using neural networks. Large-scale language models will also create extremely realistic AI companions and new social behaviors, creating never-ending games that attract and retain users for the long term. Entertainment will continue to shift from passive experience to active experience, and the boundaries between reality and virtuality will become increasingly blurred.

**Enterprise service field: Customer relationship management (CRM) driven by AI data emerges. **Generative AI will directly solve the core data problem of enterprise services. In the future, a large number of companies will no longer rely on direct interactions between sales representatives and customers, but will automatically capture from actual customer interactions (such as meeting minutes, emails and phone recordings) Or relevant data generated by AI. AI technology will enhance human creativity and problem-solving abilities and change the way we work in the future.

Finally, use ChatGPT to summarize the above content, in the new year:

**AI will become more popular and practical. **It is expected to play a role in programming, on-device AI assistants, private domain data set applications, etc., and unlock new usage scenarios.

AI will usher in technological innovation. Multimodal large models and new architectures will drive cutting-edge innovation in AI technology. Next-generation solutions such as Hyena and Mamba will emerge.

Industry applications and influence will gradually expand. AI will have a significant impact in areas such as healthcare, consumer, financial technology, gaming and enterprise services. Voice applications and professional AI assistants will shine in fields such as education and artistic creation. The integration of AI and blockchain technology will promote the global development and application of AI technology.

What these predictions have in common, ChatGPT believes:

• AI technology will continue to be deeply applied in many fields, especially in improving efficiency and creating new experiences. *Technological progress and innovation are key to promoting the development of AI, including new model architecture and multi-modal capabilities.
• The widening of industry applications and changes in business models will be driven by AI technology.
• Legal, ethical and regulatory issues will become important aspects to pay attention to in the development of AI. *The investment and financing environment may adjust as market dynamics and industry trends change.

At the end of the article, the editor would also like to thank AI for participating in this tweet and improving writing productivity.

Looking forward to the upcoming 2024, we expect that the development of AI will contribute its revolutionary power to the automation and intelligence of human social development, both at the level of basic disciplines and at the application level penetrating into all walks of life.

Reference source:

1.AI: The Coming Revolution——Coatue

2.10 AI Predictions For 2024——Rob Toews

3.Big Ideas in Tech for 2024——a16Z