How AI Changed Venture Capital

I've been involved with startups for most of my career, and investment has always been tricky. You get unicorns who instantly become profitable household names, but they're rare. Everyone has been chasing their AirBnBs and Ubers for years, usually risking massive amounts of capital. But what if you could reduce the risk to a minimum while leaving people to evaluate softer factors, leaving the heavy lifting to AI?

I recently watched an excellent video on this topic titled "How AI is Reshaping Venture Capital." It inspired most of these thoughts.

A brief aside on the history of Venture Capital

Venture capital is a form of investment in which investors invest money in startups with growth potential. This financing is provided in exchange for a share of ownership (shares) in the company.

You can roughly define its development into three stages:

1. The first phase covers the years 1948-1994. During this period, a tiny proportion of people were involved in investments. Investors and capital were limited in their capabilities. Founders mainly developed and invested in their businesses independently.
2. The second phase is 1994-2024. With the advent of the Internet, this field expanded greatly. Popular venture capital companies (Sequoia Capital, Andreessen Horowitz) and successful startups (Amazon, Facebook) appeared during this period. Over the years, venture capital has become the main driving force for the development of startups.
3. The third phase started in the 2020s. AI development has become the main aspect of scaling and expanding venture capital. It covers all industries, from robotics to biotechnology. Artificial intelligence has completely changed the approach to analysis, deal finding, and company support.

AI in Venture Capital

The emergence of AI has affected all areas, including venture capital. It has changed the decision-making process, the search for the most successful startups, and risk assessment.

1. Search and selection of startups. If earlier investors relied on personal connections, market analytics, and just intuition, now AI analyzes all risks and selects the most promising startup for investment.

2. Assessment of startup potential. AI analyzes financial indicators, the market, competition, and the team to predict the likelihood of success. Thanks to these algorithms, it takes minutes, not weeks.

3. Reduction of human bias. Thanks to this, AI helps find startups, for example, in unpopular industries that many venture capitalists ignore due to lack of expertise. AI relies on startups' digital indicators, not personal acquaintanceships or niche familiarity.

4. Management of investor assets. AI helps investors predict startup success and determine when to exit the project.

5. AI startups receive more funding. Investors invest in startups that use AI. This has led to the rapid development of the AI industry, especially in generative AI, business automation, FinTech, and MedTech.

Ethical risks of using AI in venture capital

AI has brought many effective changes to venture capital, which have expanded the scope and accelerated the process of finding and investing in startups. But the question arises:?

Can we rely entirely on AI? Let's look at the possible risks.

Complete automation of venture capital can facilitate finding and investing in investments, but it cannot wholly replace human assessment. Businessmen from this field have experience, thanks to which they can evaluate the team, their adaptability, and behavior in general. AI algorithms are unlikely to be able to catch these subtleties of human behavior and let in a project with non-standard ideas.

Bias in AI models can significantly worsen the process of finding startups. AI analyzes data from previous startups and concludes their development history. This process can lead to a limited sector of project search. For example, if historically more successful projects were founded in California, AI will conclude that projects from other regions are less relevant. This will lead to unfair selection and reduce the chances of funding successful startups.

In addition, AI uses established patterns to search for projects, so there is a risk of flooding the market with the same startups. Selecting and investing in startups that meet the AI criteria will limit the number of unique projects. For example, focusing only on projects with rapid development can deprive complex projects that require more development time (in biomechanics or medicine). This will lead to a loss of innovation in this area and a decrease in the development of scientific startups.

Therefore, AI is a tool that helps people but does not replace them. AI assessment speeds up the search and selection processes, but a person should make the final decisions with advisory inputs. People are the ones who can thoroughly analyze soft factors that algorithms cannot grasp. Much of decision-making comes down to subjective opinions that cannot be codified easily.?

Future outlook

The future in this area has already arrived. AI has already completely changed the approach to selecting and investing in startups. It's not something new we'll discuss in the future. It's something that people already do daily.

In the future, these algorithms will only improve. This will move to automated decision-making that automatically finances startups by analyzing a large amount of information, which implies the emergence of fully automated venture funds. They will significantly speed up the process of investing and signing deals, the results of which can accelerate the economy tremendously.

AI will minimize risk in this area, which can save a significant share of investors' budgets. This will allow for the correct allocation of investments and can cover many startups, bringing more significant profit to both parties.

AI will be able to give early signals of success, allowing unique startups to be found very early in their journey.

My thoughts

The main merit of AI is reducing risks. The possibility of failure scares investors and makes them biased toward startups in specific niches. This slows down the innovation process and creates inefficiencies.

AI has automated many processes to speed up the search and evaluation of startups. In a few minutes, it can process a large amount of information and spit out a list of relevant startups for financing. Machines can analyze many stories of different startups and their development to assess their business plan and predict their trajectory based on available information.

The main question remains: Is it accurate? The true answer is no, not quite yet. But it doesn't need to be 100%—it just needs to be slightly better than what we have today. And that's already been achieved.

In the future, this will further expand the market and lead to more significant investments and profits.

Much like my niche, human expertise is still key in data annotation. People are just moving from doing menial work to validating the content provided by algorithms, which allows us to do things 5-10 faster than before.