Artificial intelligence investment continues to slow down. What kind of AI projects and investment strategies can survive the cycle?

According to the "State of AI" quarterly report recently released by research organization CB Insights, consistent with the current situation in the capital market, investment in AI continues to slow down.

Total investment in AI startups fell 31% since last quarter, to the lowest level since the third quarter of 2020. Large-scale financings (more than US$100 million) fell 39% compared with the previous quarter, hitting a nine-quarter low.

Although the stagnation in AI financing will slow down the development of the field, it also prompts investors to focus more on AI projects that may achieve sustainable development. Investors need to understand the AI ​​startups that have received funding to get a general idea of ​​how the AI ​​industry will develop in the coming months.

Business Model of AI

AI startup is a vague term that generally applies to all types of companies, its scope From companies focused on providing AI tools (e.g. MLOps, predictive analytics tools, no-code/low-code model development) to companies using AI in their products (e.g. insurtech companies using machine learning to predict risk).

However, there are some factors that determine the success of business models formed around AI and machine learning. The following are some common principles of its products:

1. Product/market fit: AI products must solve unsolved problems or provide sufficient improvements on existing solutions. added value.

2. Growth strategy: There must be scalable channels for the product to deliver its value to target users (such as paid advertising and integration with existing applications). These channels must be defensive and make it difficult for competitors to capture market share.

3. Target market: Investors hope to get a return on investment. There must be a sizable market for its product to grow and reach its target valuation. If a product is too niche and few people care about it, investors won't be interested in funding it.

In addition to the above principles, products using AI and machine learning must also solve some other problems:

1. Training data: The product team needs to have enough high-quality data to train and test its model. In some cases, this data is easy to obtain (such as public datasets and existing data in enterprise databases); in others it is more difficult to obtain (such as health data). For some apps, data may differ slightly across geographies and audiences, requiring their own data collection efforts.

2. Continuous improvement: AI and machine learning models need to be constantly updated as the world changes. After deploying a machine learning model, product teams must have a strategy for continuously collecting data to update and improve the model. This continuous improvement also strengthens the product's defense against competitors.

In line with these principles, according to the CB Insights survey report, it is necessary to understand whether there is a pattern for AI startups to attract funds for their AI plans during the economic downturn.

AI projects that bucked the trend and achieved early financing

The average size of early financing in the AI ​​industry has been stable at around US$3 million. In contrast, mid- and late-stage deal sizes fell 15% and 53% quarter-on-quarter respectively. But the number of early-stage deals has shrunk, meaning AI startups will have a harder time finding investment for their product ideas.

Among the seed funding and angel deals mentioned in the CB Insights report, Israeli AI startup Voyantis received $19 million in funding in July to develop its predictive growth platform.

Today’s advertising environment has changed, with stricter regulations on user data and privacy, and Voyantis is committed to solving these problems faced by marketers. For example, Apple recently added a feature to iOS that allows users to prevent advertisers from collecting their device IDs. Without detailed data on users, previous rules-based campaigns could only deliver poor results, which would increase the cost per user acquisition (CAC). Voyantis uses machine learning to predict user behavior and lifetime value, helping to make informed decisions and improve marketing campaign ROI.

Eleven Therapeutics, another Israeli-based biotech startup, received $22 million in seed funding in August this year. It focuses on RNA therapeutics, an area that has attracted much attention in recent years, especially during the spread of the new coronavirus epidemic.

The company is developing a deep learning framework for "generating functional data on the activity distribution of siRNA molecules." There isn't much information about the company's AI technology, but it's a market space with plenty of potential, and its financial backers include the Bill & Melinda Gates Foundation.

US-based startup Spice AI received $14 million in seed funding in September this year and is building digital infrastructure for creating AI-driven Web3 applications. Interestingly, the company managed to attract investment at a time when the crypto startup industry was in worse shape than other industries.

There are three things worth noting about this company: First, it is creating data engineering infrastructure to index existing data on major blockchains, which means it does not have any major obstacles in getting the data. Second, its founders are Microsoft Azure veterans, including Chief Technology Officer Mark Russinovich and the former and current CEO of GitHub (acquired by Microsoft in 2018). Having such a high-profile industry figure makes it easier for the company to attract investment, even in the most difficult of times. Third, blockchain data engineering is largely an unsolved problem that Web3 companies will certainly face as the industry matures, so this can be considered one of Web3's lower-risk projects.

Who has received huge investments in the field of AI?

Among the startups that received huge financing in the third quarter of 2022, the American startup Afresh received raised US$115 million in Series B financing. The company uses machine learning to help grocery store operators reduce food waste by up to 25%, as the platform tracks fresh food sales and helps predict future customer demand. Supply chain teams can use the platform to optimize procurement, and users can place orders directly with suppliers using the platform to reduce food waste.

The company already has thousands of customers in 40 states in the United States and will use the new financing to grow its business, expand its market to other countries and regions, and add new features to increase the value and value of its products. Market coverage.

Another company that received huge investment is Italy-based mobile app developer Bending Spoons, which raised $340 million in September this year. Bending Spoons develops mobile video and photo editing apps that use machine learning to perform complex tasks such as background removal, automatic captioning, and photo enhancement.

The company's application adopts a freemium model, where users can use basic functions for free, but must pay to use advanced functions. Founded in 2013, Bending Spoons has been downloaded more than 500 million times and has annual revenue exceeding $100 million for several years. The next step will be to use the new financing to develop new products and make acquisitions, market its new products to existing customers, and Collect more data to further expand your lead over your competitors.

AI investment rules through cycles

If you delve into the AI ​​companies that have received financing, you will get more information, but pay attention to the following points:

1. Adhere to good product principles: No matter how good the AI ​​is, it needs a product that can solve real problems. It is much better than other products and has less resistance to adoption. At the same time, AI products also need to have a huge market, room for expansion, and a clear vision for sustainable growth.

2. B2B AI is the most important: While AI-driven applications provide convenience to consumers, their value to businesses is much greater, especially as the economy enters a recession in the case of. Well-implemented AI can reduce wasted money, optimize recommendations, and automate manual functions, all of which impact an AI company's expenses and revenue.

3. Find new AI markets among unsolved problems: In the field of AI, established markets are difficult to conquer because existing AI companies already have better dataset to train their model. And it’s easier and cheaper to enter new markets, especially if you can quickly collect data to train machine learning models before your competitors do.

4. Reduce the cost of acquiring data: Look for AI ideas where the data already exists and is annotated (e.g., financial transactions, sales history, patient records). Or look for solutions that generate the data needed for the model to reduce the need for data collection. If an enterprise's application requires a new pipeline to collect, clean and annotate data, it will require more time, talent and money, which is difficult to achieve in the current situation.

5. Having well-known founders will attract more investment: Founders who have worked in large technology companies are more likely to attract AI companies (such as Web3AI’s data infrastructure) More and investment.

The above is the detailed content of Artificial intelligence investment continues to slow down. What kind of AI projects and investment strategies can survive the cycle?. For more information, please follow other related articles on the PHP Chinese website!