Research

The AI sector is experiencing exponential advances in technology with a new revolutionary developments almost daily. Investment in the tech sector is down significantly, but investment in the AI sector has only decreased slightly by comparison. M&A in the sector is very active with large technology companies the most acquisitive. Valuations are astronomical, which is typical for a large new exciting sector, but coming down to earth slowly. In this dynamic environment, there are practical steps founders and investors can take now to capitalize on the AI sector.

I. Revolutionary Technology
Below are three examples of how AI will revolutionize the world:
• DeepMind’s AlphaFold model helps predict the structures of 200 million proteins and is being used by more than one million biologists around the world.
• Cognition’s AI coding tool, Devin, can execute complex software engineering tasks such creating software from a single prompt which will automate software development.
• OpenAI’s Sora is an AI model that can create realistic and imaginative scenes from text instructions. Here is Sora’s first music video where it was director and scriptwriter: Worldweight.
The industry map below by the noted venture capitalist Matt Turck shows just how complex the sector has grown:

Figure 1: The 2024 MAD (Machine Learning, Artificial Intelligence & Data) Landscape

https://mattturck.com/landscape/mad2024.pdf

II. Investment
Venture funding in the AI sector topped $50 billion in 2023. In Q1 2024, global startup funding was the second-lowest on record since the beginning of 2018 but the AI sector raised $11.4 billion, only a slight decrease from 2023 on a quarterly basis of $12.5 billion. The AI sector came in first for funding in 2023 but in Q1 2024 the AI sector came in second after healthcare (Crunchbase).
There are very ambitious projects on the horizon requiring massive investment with Microsoft and OpenAI planning to build a $100 Billion “Stargate” AI supercomputer and Sam Altman raising up to $7 trillion for his new AI chip project.
However, investors are scrutinizing AI companies more to understand if the companies are real AI or just have an API to an LLM. How many engineers on staff are truly AI experienced? Is there a product market fit with growing revenue generation? When 57% of the latest Y Combinator batch W24 are building applications or tools around artificial intelligence and many companies are rebranding themselves as AI companies, investors are digging deeper.

III. M&A
Companies are rushing to acquire AI startups and resources to boost their AI capabilities across various sectors like cybersecurity, enterprise software, semiconductors, biotech, and legal tech. Private equity firms, sitting on over $2 trillion in dry powder, are placing bets on AI gold rush with billions pumped into datacenters (S&P Global). The large tech companies are acquiring companies rapidly. Apple acquired a total of 32 AI startups in 2023, securing its lead in the AI sector among major tech companies.

Figure 2: 2023 Acquisitions by Large Tech Companies (CDO Magazine)

 

https://www.cdomagazine.tech/aiml/ai-shopping-spree-apple-leads-charge-with-32-startup-acquisitions-in-2023
IV. Valuation
Valuations are astronomical, but coming down to earth slowly. Hugging Face, an AI infrastructure platform focused on open-source development, has one of the highest revenue multiples at 150x ($30M in 2023 revenue at a $4.5B valuation). It’s followed by Perplexity, which is developing an alternative to traditional search engines, at 65x (based on a 2023 valuation of $520M and $8M in 2024 ARR). Perplexity AI’s revenue multiple dropped to 65 from 150 between its last two funding rounds, which may be a sign that valuations in the AI sector may finally be coming down to earth (The Information).

Figure 3: Revenue Multiple by Company (CB Insights)

https://research-assets.cbinsights.com/2024/03/25093112/AI-100_-Revenue-multiple-by-company.png
V. Practical Tips for Founders and Investors
Position Your AI Company
In order to attract investors and acquirers, an AI company needs to be positioned correctly. There are several categories:
AI Infrastructure: These companies develop technology that improves parts of the AI creation or deployment process itself, such as data cleaning, modeling, or deployment.
Horizontal AI: These companies develop AI-powered technology to help businesses perform specific tasks across different industries.
Vertical AI: These companies apply AI techniques to solve specific business problems in particular industries. Examples include AI for healthcare, finance, and manufacturing. Most AI companies fall into this category.

Figure 4: The AI 100 (CB Insights)

https://www.cbinsights.com/research/report/artificial-intelligence-top-startups-2024/
Established companies with proprietary databases are competing with AI startups, for example, Johnson & Johnson has an AI research department J&J Innovation. Countries already have national champion LLMs, e.g. Open AI in the US, Cohere in Canada, Mistral in France and Moonshot in China. Moonshot just raised $1 billion at a $2.5 billion valuation (TechCrunch).

Lock up Engineers

There is a heated talent war for AI engineers. The best engineers easily get total compensation packages of $1 million a year (Wall Street Journal). Microsoft hired the co-founders and the whole team of engineers at the start-up Inflection (Financial Times). Lock up your engineers with generous salaries and equity packages and, if possible, have them sign employment agreements with non-competes.

Cast a Wide Net
Investors should consider traditional venture capital firms as well as corporate venture capital, e.g. the Open AI Startup Fund. For M&A, large tech companies such as Microsoft, Nvidia and Apple should be on any list. Other large tech companies should also be considered, e.g. Databricks acquired MosaicML for $1.3 billion to add capabilities for business customers to build their own AI models. Vertical AI companies should consider non-AI companies in the same vertical seeking AI capabilities, e.g. Thomson Reuters acquired Casetext for $650m for its AI legal research technology.

VI. Conclusion
Bill Joy in the article “Why the Future Doesn’t Need Us” published in Wired Magazine in 2000 predicted that “eventually a stage may be reached at which the decisions necessary to keep the system running will be so complex that human beings will be incapable of making them intelligently.” In the meantime, before AI is smarter than humans, founders and investors can make intelligent decisions now about capital raises and M&A to capitalize on the AI sector before Hal no longer opens the pod bay doors.