A.I. frenzy from Alphabet hits the startup scene with CapitalG investing $100 million in AlphaSense

Google has made a significant investment in artificial intelligence and is at the forefront of the latest boom in generative AI, but its investments don't end with its homegrown products. There is also a lot of money being invested by Alphabet's parent company in the startup industry.

‍

CapitalG, Alphabet's late-stage venture capital arm, told Trade Algo that it has just led the $100 million investment in corporate data firm AlphaSense, valued at $1.8 billion, which is the biggest investment the company has ever made. The company competes with companies like FactSet and Bloomberg in terms of providing information about businesses that can be used by corporations and investors to make informed decisions.

‍

An announcement of the funding round was made on Tuesday evening after months of increasing hype surrounding generative artificial intelligence, particularly OpenAI's ChatGPT and other text-generating tools that use large language models or LLMs to provide creative and sophisticated answers to user queries. Google introduced a conversational technology called Bard in February that will work with its dominant search engine, as well as other products from the company.

‍

AlphaSense's CEO Jack Kokko recently announced that the company is working on a new feature that will automatically summarise financial documents for its customers so they will be able to get a better understanding of their finances. LLMs have significantly improved the summarization process with the help of artificial intelligence software, which has historically been a challenging task for AI programs.

‍

Kokko is the founder and CEO of AlphaSense, which was founded in 2011. There has been a flat round of financing following a $225 million investment made by Goldman Sachs and Viking Global Investors in June of this year. In that round of funding, the valuation of the company doubled from what it was in a 2021 round that was led by the same investors.

‍

During the last two rounds, the topic of generational artificial intelligence was not a hot topic as the term had not yet entered the popular lexicon. According to a press release issued by AlphaSense last year, the platform's “proprietary search technology” is powered by artificial intelligence (AI) and natural language processing and is able to “extract relevant insights from an extensive library of public and private content.”

‍

In the wake of a plunge in public company valuations and a freeze on the IPO market for the first time since 2021, the tech funding market has dried up. It is hard to deny that generative AI has turned into a rather frothy subject in some corners this year. It was announced in March that a 2-year-old pre-revenue startup called Character.AI, founded by two former Google employees, had raised $150 million in a pre-revenue round led by Andreessen Horowitz at a valuation of $1 billion.

‍

With AlphaSense, we are much further along than we were before, having already been able to surpass $100 million in annual recurring revenue in 2022. In order to prepare for the company's IPO, Kokko said the new capital will be used to hire additional salespeople as the company prepares to go public when the economy stabilizes. The money will also help AlphaSense improve its technology, utilizing advances in generative artificial intelligence (AI) to maximize the use of the company's technology.

‍

As James Luo, a CapitalG partner, pointed out, part of the appeal of AlphaSense using newer LLMs is that it can make the core product more appealing to customers outside of the traditional financial service industry and make it more user-friendly. It is quite likely that salespeople will use a product such as AlphaSense if its interface is more intuitive and easy to use, for instance.

‍

"There are a lot of people using Google Search to try to find every piece of information they can find, but they don't have access to a lot of proprietary content," Luo said, referring to the potential new users. "As someone who is not working in that world, it's important for you to have something that will help you understand this information a lot better."

‍

However, modern-day LLMs suffer from a phenomenon that AI researchers refer to as "hallucination," in which the software is prone to generate inaccurate responses when it is asked a question.

‍

In addition, Kokko said AlphaSense is working on a variety of techniques to ensure that its technology is able to provide accurate summaries along with footnotes to documents so that people will be able to identify the sources of information. Although he declined to identify the specific LLMs which AlphaSense intends to incorporate, he said the company is currently testing nearly every major model that is currently available to it. Alphabet and Meta are examples of tech companies that have developed LLMs, as are startups such as OpenAI and Cohere.