AI Art Adviser

Fine art is an investment vehicle with an average annual return of 6.5%, according to Stanford-based economists. However, traditional art advisers who help assess fine art returns tend to only work with individuals with a seven-figure or more yearly fine art budget.

We are working with Stanford's BlackAIR on an artificial intelligence-based research project to help lower the cost of predicting average returns. Traditional price index methodologies reinforce systemic biases because they rely too heavily on past recorded transactions. Research shows that 98.8% of past recorded transactions are of artwork by non-black American artists. Research shows similar disadvantages among women artists and other minority artists. Our model doesn't rely solely on past recorded transactions and it's attempting to avoid these biases.

The solution could not only help more individuals have access to fine art for the investment potential, but could help more BIPOC and women artists sell their work.

We asked ourselves the question, "Why don't the middle class invest in original fine art? Particularly, if research shows it grows faster than bonds?" As we started to research the answer, we found that the traditional art buying class of white males might be influencing the types of artists who are able to make a living as full-time artists.

If we could expand the types of people buying art, our research shows that we could directly influence the types of people who are able to make fine art full-time.

We developed a patent-pending system that educates the general public on fine art for free through mobile phones. Anyone can use the app and "become a collector" with a defined aesthetic preference, such as a preference for 17th century Japanese porcelain or a preference for Cuban cubists. Our artificial intelligence informs everyone of their preferences through the rating system.

We then use their ratings to create pop-up galleries closer to users most likely to convert to customers. We found that when potential (or current) collectors learn about suspected fine art appreciation, they are most likely to buy. So we are also developing a system to predict fine art appreciation that doesn't include typical biases of the standard fine art price indexes.

Giving this information via artificial intelligence (AI) allows us to give the general public access to something that is usually only reserved for wealthy consumers. In addition, by using AI and collecting data from a wider variety of people than traditional fine art pricing analysts, we believe our model will have ingrained in it less systemic biases.

Our solution serves both the artists and the middle class. For the artist, we believe that AI can predict buying patterns, and we hope to develop a system that allows more artists to make a living wage each year. We will also continue developing a system that acknowledges institutional biases within the fine art industry and uses technology to address those issues fairly.

If one were to buy Andy Warhol in the 1970s, it would have been about USD 25,000 in today's dollars. However, today his pieces regularly sell for multimillion dollars. We believe that creating a system that encourages more consumers to have access and feel informed about fine art's investment potential could lead to more wealth in the hands of the general consumer.

Our team has conducted quantitative and qualitative surveys with artists to develop a system that is both easy to use and would encourage increased average spend. We are now in the process of conducting interviews also to incorporate and support fine art museum development. In a National Science Foundation-supported study, we are now conducting interviews to find out how our system can increase the diversity of their donors. We believe that greater diversity in who supports museum funding will directly impact the types of works supported by museums.

The very wealthy hold a variety of financial assets, including real estate, stocks, bonds, gold, and fine art. Several banks allow fine art to be used as collateral for personal loans. It's a financial instrument that can both provide credit and generational wealth, but it has not traditionally been used in communities of color. We have a designed system that encourages increase consumption of fine art and increased education of fine art's investment potential.

We have built a prototype, and we are in the process of testing it this summer with the help of Stanford University Black in AI's Innovation and Research summer grant. In addition, we are running a test on whether our predictor model of fine art appreciation will be more inclusive and increase the consumption of BIPOC artists than traditional pricing models.

We are also in the concept stage. We are working with Midwest i-Corps to develop a new feature to help museums identify high-net-worth individuals who have an interest in their collections. We believe that this relationship could help museums also identify more BIPOC donors who may positively impact curatorial decision-making.

There is a lot of talk about artificial intelligence (AI) from the standpoint of bias, but AI can also lower the cost of services or goods. Our team developed this solution because we are aware that traditional art advisers tend to only work with people who have a seven-figure or higher art budget because of how they generally price their services. Traditional art advisers tend to add a five to ten percent fee on top of all the art that you select to buy.

We  have iterated multiple times to find a way to serve a wider audience because we believe that more people buying art will lead to more support of diversity. The AI can also help us find patterns of interest. Typical galleries tend to have two or three customers that buy the majority of the work each year. This causes gallery owners to have a thesis or a vision of what to sell that may not correlate to the data of what people actually want. Our model supports the public making their own choices in art and not being dictated by others.

We also plan to use the data to support the creation of pop-up galleries in underserved areas outside of the major metropolitan areas. Our research shows that there is a hunger to be served in high-income areas outside of the traditional art scenes of New York City or Los Angeles. We believe that connecting these areas to the best fine art in the world could have the potential to accelerate wealth creation to both artists and collectors.

We have a waiting list of 203 people. We would like to accelerate our growth to 10,000 users by the end of the year. Artists on our waiting list have indicated that they would share the app with 70 people on average. In the next five years, we would hope to saturate the market at 100 million users and to have the app available in multiple languages.

We are now in the process of testing our method of predicting financial return against the traditional methods. We are under the assumption that our method will be the only model able to include more women and minorities, but our research will not be completed until September 2021.

In the future, we would also like to test whether introducing museums to lower-level wealthy individuals would increase the diversity of the donors. We would also like to study how much this would possibly influence the diversity of major museum collections.

1 Full-Time Founder

1 Part-Time Founder

20 Contractors

The founding team is my sister and me. When we were growing up in all-white communities, my parents had a pair of paintings of an African couple that hung on the walls in our foyer. Oftentimes, I experienced racism in these all-white communities, but those paintings staring back at me as I entered the house allowed me to reposition my identity back to that African couple. As I grew up, I realized I needed to decorate my own home with those same standards, but I found the art market unnecessarily cumbersome, it took too much time to find what I wanted and I also didn't have the necessary resources to make that type of financial commitment.

My younger sister, Anya, was just learning about artificial intelligence in college. So when I told her how I was trying to lower the cost of art advising, she said AI would be better and I could probably lower the cost to free. In three months, we built out a prototype. My former background was in survey research and data analysis, so I tested the theory with both qualitative and quantitative surveys during the app development to continue to validate the concept.

We have several leadership advisers who are a mix of men and women of various backgrounds. Both co-founders are women of Afro-Latina backgrounds. For our research studies, we conduct purposeful sampling, by reaching out to diverse communities to maintain an accurate demographic sample. If our company is able to hire other full-time workers, we believe that we would attempt to hire purposefully to maintain a level of diversity that is representative of the country that we are working in.

The Solve grant would help both co-founders complete the summer research study. It will also help our research gain exposure by the media and at conferences. We hope to highlight how technology can be used to solve institutional bias and increase financial stability for non-traditional careers.

• MIT D. Fox Harrell because of his research background on the intersection of artificial intelligence and visual arts.
• Elevate Prize Foundation Caroline Garcia Jayaram because of her background in fine arts.
• UCLA Stephen Nelson because of the abundance of research on fine art economics that he has produced during his academic career.
• Museums interested in reaching out and finding more diverse donors and committing to more diversity in their collections.

If we had access to that amount of capital and level of support, we could accelerate our growth timetable. We would be able to complete the necessary app updates for open beta testing by December and accelerate our ad spend. To hire the extra developers required to complete the next round of updates is about $40,000. Currently, it costs us between $1 to $1.50 to acquire a waitlist e-mail. If given this capital, we could complete the updates and complete this year's goal of reaching 60,000 users.

We could also accelerate our timeline for the next app update on helping to encourage museums to find more diverse donors. Initial testing with the National Science Foundation-supported study has been positive, but the funds could be used to complete the prototype.