Part 2 AI: Easy to Discuss the Hype, But What About Substance

Artificial Intelligence

With a topic as exciting as Artificial Intelligence (AI), it’s easy to get wrapped up in the hype. Gartner has even delineated its ‘hype cycle’ to more or less codify the emotional journey that frequently accompanies advancements in technology1. It’s easy to see a new achievement and think of how exciting it is, but in thinking about an investment thesis, it’s important to always step back from the emotion and consider where the rubber really meets the road on such topics.

Anthem: What does ‘doing AI’ actually look like?2

Anthem Inc. is a health insurance company, it was recently reported that they would be working with Google Cloud to generate 1.5 to 2 petabytes of synthetic data.

 

Pausing for a moment—what is a ‘petabyte’? We are much more familiar with ‘gigabytes’, where a two-hour movie might take up 2-4 gigabytes depending on the video quality used. 1 terabyte is roughly 1,000 gigabytes and then 1 petabyte is 1,000 terabytes. So, when we think of a petabyte it is 1 million gigabytes, the unit with which we are more familiar.

 

Meta Platforms is building a research super cluster focused on training models on an exabyte of data—which is the next step up and 1,000 petabytes. 1 exabyte is roughly equal to 36,000 years of high-quality video3.

 

Secondly, what is ‘synthetic data’? It could be real-world data that has been stripped of personal information and fully anonymized, or it could be completely artificial, generated from deep generative models. An interesting side benefit, at least possibly, in using synthetic data, is in how it could mitigate certain biases that have been shown to exist in real world datasets.

 

So, why does Anthem need to generate 1.5 to 2 petabytes of synthetic data—what problems could that help in solving? The company says that the data will be used to validate and train AI algorithms that identify things like fraudulent claims or abnormalities in a person’s health records. Anthem uses both Amazon Web Services (AWS) and Google Cloud for cloud computing. The company decided to use Google’s services because of their experience in using AI technology.

Black Knight: Bringing AI to mortgage data

On 4 May 2022, Intercontinental Exchange Inc. (ICE) agreed to buy mortgage-data firm Black Knight in a deal valued at $13.1 billion. It is expected to close in the first half of 2023. ICE is known as an operator of exchanges, clearinghouses and other financial market infrastructure, but in recent years it has pivoted towards the growing digitization of the housing market. Prior to the Black Knight deal, ICE had purchased the mortgage-software firm Ellie Mae for $11 billion, and it had also purchased Simplifile, a firm that facilitates the electronic processing of mortgage records4.

 

Black Knight itself started incorporating AI as a value-added service to their solutions in 2018. It was noticeable seeing Artificial Intelligence Virtual Artist  (AIVA), a mortgage AI digital assistant being used, which was done via the acquisition of HeavyWater in 2018. The company was also very acquisitive in the AI space with collateral analytics (automated AI-based real estate data analytics), engineering in medicine and biology society (eMBS) (cloud-based analytics platform), Top of Mind (AI-based customer relationship management (CRM) software) and Optimal Blue (automated loan origination platform). Software solutions represented more than 80% of Black Night’s revenue at the close of 20215.

Synopsys: Part of the foundation for ‘smart everything’

Synopsys posted a second quarter profit of $294.8 million, which topped analyst expectations. The market responded particularly well to the firm increasing full-year revenue guidance to a range of $5.0 to $5.05 billion, up from a prior target of $4.78 billion to $4.83 billion6.

 

Synopsys is an electronic design automation company, and it is particularly strong in field effect transistor design, which could provide growth opportunities. Customers of Synopsys include most major semiconductor companies. If one is thinking of the market opportunity for Synopsys, one must consider the increasing demand for ‘smart everything.’ In their investor presentation, Synopsys included a forecast that deep learning chipset revenues could reach $72 billion by 2025, from a level of less than $10 billion in 20187.

Splunk: Resilience in the face of economic turmoil

Splunk delivered stronger results than the market community was expecting when it reported its results for the quarter ended 30 April 2022. Revenue increased to $674 million, a 34% increase over the prior year. Cloud revenue was $323 million, up 66% year-over-year for the specific segment. Company leadership indicated that Splunk is not at risk of too much impact from current macroeconomic issues because company security budgets do not appear at risk8.

Silicon Motion: Being acquired by MaxLinear

In early May 2022, it was announced that the Taiwan semiconductor firm Silicon Motion Technology agreed to be acquired by MaxLinear in a deal valued at $3.8 billion9. Silicon Motion has over 20 years of experience developing specialised processor integrated circuits that deliver market-leading storage solutions that are used in data centres, personal computers, smart phones and in commercial and industrial applications. The portfolio of controller intellectual properties is extremely broad10.

Palo Alto Networks: AI & machine learning could add value across entire platform

The Russia/Ukraine conflict has kept a strong focus on cybersecurity in 2022, helping Palo Alto’s results. Similar to Splunk, the company have not yet seen the heightened inflation causing pressure on company’s cybersecurity budgets. As Palo Alto Networks lay out their view of the market landscape, they clearly indicate that most companies want greater security.

Sources

 

1 Source: https://www.gartner.co.uk/en/methodologies/gartner-hype-cycle

2 Source: Bousquette, Isabelle. “Anthem Looks to Fuel AI Efforts with Petabytes of Synthetic Data.” Wall Street Journal. 17 May 2022.

3 Source: Bhattacharyya, Suman. “Meta Unveils New AI Supercomputer.” Wall Street Journal. 24 January 2022.

4 Source: Osipovich, Alexander. “Intercontinental Exchange to Buy Mortgage-Data Firm Black Knight for $13.1 Billion.” Wall Street Journal. 4 May 2022.

5 Source: Consumer Technology Association

6 Source: Moore, Logan. “Synopsys Stock Is Rising on Strong Review. Analysts See Growth Opportunities.” Barron’s. 19 May 2022.

7 Source: https://www.synopsys.com/content/dam/synopsys/company/investor-relations/corporate-overview-investor-q2-2022-final.pdf

8 Source: Savitz, Eric J. “Splunk Stock Rallies as Software Company’s Results Top Estimates.” Barron’s. 25 May 2022

9 Source: Saigoal, Lina. “Silicon Motion Technology Stock Soars After $3.8 Billion MaxLinear Deal.” Barron’s. 5 May 2022

10 Source: https://siliconmotiontechnologycorporation.gcs-web.com/static-files/7708fce8-a220-49b6-8c11-5dd10b776f5f

Leave a Reply

Your email address will not be published.