If you ask ChatGPT what the most exciting trend for AI in 2023 will be, the bot answers: “explainability and transparency.” At the same time, CNN recently released an article calling 2023 the year of generative AI. The writer also joked that “resistance” to AI’s new market dominance is “futile.” While this sentiment might seem dramatic, people really are racing to adopt generative AI tools. At the intersection of the increased interest in adopting AI and the focus on increasing the transparency of AI, sits the biggest areas for growth in the AI market. Given all the hype, we thought exploring where Summari sits in the AI market would be useful. For starters, Summari is a generative AI company. 

The generative AI market consists of three layers of companies: 

1. Infrastructure vendors; e.g., Microsoft Azure, Amazon Web Services, Google Cloud‍
2. Model providers; e.g., Cohere, OpenAI, Anthropic‍
3. Application companies; e.g., Jasper, Summari 

Infrastructure vendors run cloud platforms and hardware that train large-scale models and host workloads for generative AI models. Model providers allow access to the models that power AI products through APIs or open-source platforms. Application companies integrate models into user-facing products using their own model pipelines and/or a third-party API. Summari is an example of an application company.

According to Andreessen Horowitz, a venture capital firm that backs technology companies, Application companies have a lot of room to grow and much to gain in the generative AI market. In order to grow, they must build defensible moats or risk being copied and outmaneuvered as new products launch. Model providers, the backbone of the generative AI market, have yet to achieve significant commercial-scale success, but are promising. A recent report from Grand View Research projects that the size of the global generative AI market will reach $109.4 billion by 2030. According to the report, the fact that the quality of AI products has increased exponentially in the last year has helped drive the adoption of AI solutions. Factors like better image resolution in image generators, reduced latency, enhanced performance, and the speed of outputs are driving the market value of generative AI products. 

CNN predicts that 2023 will be the year of generative AI, and the adoption of trustworthy tools will be paramount to the success of application companies. One example of the importance of product’s being trustworthy is the recent debacle with Google’s search engine chatbot Bard. While huge announcements from Google sent its stocks sky-high as it entered the arena to contend with Microsoft, the speed with which its generative AI product was released came back to hurt Google, when in a demonstration posted to Twitter, Google’s new AI assistant answered a question incorrectly. Bard answered, “what new discoveries from the James Webb Space Telescope can I tell my 9 year old about,” with an image taken by the European Southern Observatory’s Very Large Telescope in 2004. When the error was pointed out by Reuters, Google shares fell 7.7%, taking $100 billion off its market value. 

The need to be cautious about the roll out of AI products sparked interest in the MIT spinoff Verta, which is looking to find a niche between the pressure to adopt AI and the need to monitor and maintain AI models. Verta provides a platform to help companies monitor, manage and deploy the machine-learning models they use to ensure they are safe and performing as expected in the real world. Model monitoring is a huge opportunity for growth; Verta is already working with big players in healthcare, finance, and insurance. The company is catching the eyes of AI adopters and investors for its ability to accelerate the machine learning pipeline while maintaining the integrity of the model. Verta is an example of a company looking to capitalize on the opportunities for growth in the generative AI market.

For 2023 to truly be the year of generative AI, AI tools must transition from cool demos to reliable production-ready applications. As demonstrated by Google, and the success of Verta,  the stakes are incredibly high, especially for the early adopters of AI tools. According to the IBM AI Adoption Index Report from 2022, 35% of companies reported using AI in their business, and an additional 42% reported they are exploring AI. With all the exciting adoption, many companies haven’t taken the key steps to vet the AI tools they are using. According to the report, 74% weren’t working towards reducing unintended bias in their use of AI, 68% weren’t tracking performance variations and model drift, and 61% were not doing enough to make sure they could explain the decisions they made with input from AI. In addition, 85% of IT professionals surveyed agreed that consumers were more likely to decide on a company that is transparent about how its AI models are built, managed, and used. 

As evidenced by the survey, there is an immense drive to adopt AI but a lack of resources to ensure that AI solutions are trustworthy. As an application company that is focused on providing a trustworthy product, Summari occupies a niche in the generative AI market as an application company devoted to delivering a trustworthy product and committed to transparency. Transparency is a critical component of establishing trust when using AI, and the trustworthiness of an AI solution can easily be forgotten when solutions are adopted rapidly. Coupled with the growing trend of investing in AI companies, the generative AI market’s future will require companies to lean into explainability and transparency. Companies like Summari or Verta who are finding a niche in the useful, trustworthy generative AI space have the most to gain.