Artificial Intelligence & Machine Learning Insights

By Mark Arnold (Hyperion CIO) and Jason Orthman (Hyperion Deputy CIO)

The Hyperion Global Growth Companies Fund (Managed Fund)¹ is well positioned to benefit from any inflection point from artificial intelligence/machine learning (AI/ML) usage. The release of ChatGPT has created meaningful hype, but many of the broader use cases are real and embedded within market-leading companies. Many of our portfolio holdings have employed AI/ML technologies over the past 10 years. These dominant technology-based companies are well-placed to capitalise on this AI/ML trend. Thus, we do not need to speculate on emerging companies with nascent technologies.

AI tools, such as large-language models (LLMs), are likely to become commoditised over time, whereas proprietary datasets that companies control and train these models on will not be. Those corporations with the largest and most high-quality data will win, including Platforms with proprietary, first-party (1P) data and high usage rates. AI models trained on these data sets will produce superior results.

The majority of the Hyperion Global Strategy holdings should benefit from advancements in AI/ML, including Microsoft, Tesla, Meta, Amazon, ServiceNow, Palantir, Workday, Salesforce, Spotify, Block, Intuit, Alphabet, and Airbnb. In our view, the benefits far outweigh any disruption risk across the portfolio.

We believe there are three key benefits from AI/ML usage:

1) Productivity benefits. This includes automating processes such as product development and go to market. It should be a meaningful internal efficiency tool for these organisations.
2) Incremental revenue opportunity. We expect management teams will engage with customers in order to monetise AI/ML capabilities. They will price these services through additional pricing tiers, modules or subscriptions. For example, Microsoft is currently trialling this.
3) Strengthen their sustainable competitive advantages. The already dominant platforms will have more and better data to train their models, resulting in a better offering than their smaller competitors. This should create a virtuous loop of more customers being enticed to the stronger offering leading to greater volumes of data created for training large AI models and a more compelling offering.

We believe the portfolio is well placed to benefit from advancements in AI/ML and it should be a material tailwind over the medium term.

• Tesla Inc.: has made significant advancements in AI/ML over the past decade through its investments in Full Self-Driving software, Neural Networks, Dojo supercomputers, and the Tesla Bot. The company has significant advantage through its development of both AI software and hardware, including its fleet of over 4.5 million vehicles which generate significant volumes of real-world driving data.
• ServiceNow Inc.: has introduced a number of generative AI features, including providing connectors to existing LLMs and through developing domain-specific LLMs (i.e. for ITSM, CSM, etc.) that will be trained on ServiceNow-specific data.
• Amazon Inc.: has embraced AI/ML across all aspects of its business to improve operations and enhance the customer experience. These include personal product recommendation algorithms, Alexa-enabled voice shopping, supply chain optimisation, fraud detection and customer service.
• Microsoft Corporation: Microsoft AI, powered by Azure, provides billions of intelligent experiences every day in Windows, Xbox, Microsoft 365, Teams, Azure AI, Power Platform, Dynamics 365 and Microsoft Defender.
• Workday, Inc.: AI/ML have been embedded in Workday’s HR and Financials offerings for almost a decade, and leverages 1P data (in a multi-tenant cloud) from its over 60 million users that create more than 600 billion transactions per year. Workday’s scale is difficult to replicate, creates a virtuous cycle, and offers distinct advantages to its clients.
• Spotify Technology SA: has introduced a new AI DJ that will further personalise the audio experience for Spotify users. Spotify is also investing in creating an AI voice platform with the potential to lower hurdles to creating audio-specific content.
• Block, Inc.: plans to deploy 30 applications for generative AI in its Square ecosystem, helping its merchant customers gain greater insight from data generated from their payments system.
• Intuit Inc.: Intuit’s proprietary generative AI operating system (GenOS) and custom-trained financial LLMs generate 58 billion ML predictions per day (730 million AI-driven customer interactions per year) and specialise in solving tax, accounting, marketing, cash flow, and personal finance challenges for its 100m+ customers around the world.
• Intuitive Surgical, Inc.: With over 20 years of usage and more than 10 million procedures performed within a closed system, Intuitive Surgical is using AI/ML technologies to deliver insights to surgeons that increase the consistency of patient outcomes.
• Airbnb, Inc.: has been incorporating AI into the centre of its app by building on available LLM infrastructure to provide the best-in-class multi-modal interface with finetuned personalised recommendations and AI-augmented customer service (improved speed, accuracy, and consistency). The role of Airbnb is to be the ultimate host.
• Palantir Technologies Inc.: core offerings have always revolved around the use of software and AI/ML to help governments and large enterprises make better use of their structured and unstructured data from disparate data bases. Palantir’s new AI Platform specifically aims to address privacy/security concerns in the adoption of LLMs by applying industry-leading guardrails to define permissions and controls.
• Salesforce, Inc.: Salesforce has mobilised its entire company around the generative AI opportunity through GPT integrations for its cloud products (Sales GPT, Service GPT, Slack GPT), AI Cloud as an enterprise focused bundle of core CRM products (real-time data, analytics, and automation across applications and workflows), Einstein GPT Trust layer (industry standard for AI data privacy and security), and generative use cases in sales, marketing, commerce and for developers.
• Alphabet Inc.: has utilised AI/ML functions for many years, targeted at offering users more personalised and relevant experiences, increasing engagement and usage of its existing products. This year, the company has released its AI chatbot, Bard, and its Google AI search enhancement, Search Generative Experience (SGE).
• Meta Inc.: has incorporated AI/ML solutions across the platform through the development of advanced algorithms that rank feeds and search results, creating new text-understanding algorithms that keep spam and misleading content at bay, and automatically caption videos through speech-recognition systems. Building on Meta AI’s key principles of openness, collaboration, excellence, and scale, the company is focused on pushing the boundaries of AI to create a more connected world.

Case Studies

Workday:

AI and ML have been integrated into Workday’s operations for almost a decade and identifies tangible ways to extract value from LLMs. The company’s AI and ML models are embedded in its HR and Financials offerings which leverage a unified data core, in a normalised multi-tenant cloud, with data from over 60 million users representing more than 600 billion transactions last year. Additional to extensive 1P data, Workdays Prism Analytics can co-mingle 3P data. Workday’s scale is difficult to replicate, creates a virtuous cycle, and offers distinct advantages to its clients. Workday is committed to further strengthening its prowess in this field, as evidenced by a recent $250 million expansion of its venture capital fund dedicated to driving innovation in AI and ML. Examples of AI/ML use cases include:

• Workday’s Skill Cloud is used by nearly 50% of all live Workday HCM customers and embedded with AI and ML that empowers customers to make data-driven decisions around recruiting and the skills and capabilities of their existing workforce.

• Workday Peakon Employee Voice provides AI summation and meaningful insights from millions of employee comments to drive positive business outcomes.

• Predictive forecasting provides AI/ML-based forecasts, regressions, and predictive plans for the office of the CFO, helping forecast business demand and improve cost effectiveness.

• Workday Extend APIs enable customers to build extensions that leverage ML.

Tesla:

Hyperion has always held the view that Tesla Inc. is more than an automotive company and a significant part of its value lies in the cutting-edge AI hardware and software that it develops and deploys. Tesla has made significant advancements in AI/ML over the past decade through its investments in Full Self-Driving software (FSD), Neural Networks, Dojo supercomputers, and the Tesla Bot. Compared to its competitors, Tesla has a unique advantage in that it has a fleet of over 4.5 million vehicles feeding real-world driving data into its AI model. The company is then able to train and improve its models extremely quickly before implementing upgrades across its entire fleet via over-the-air updates. Additionally, Tesla’s capabilities in AI would enable the Tesla Network (an Uber-like ride-sharing network of fully autonomous vehicles) which would revolutionise the global transport market and represent one of the greatest organic growth drivers ever seen. The same technology is also currently being integrated into the Tesla Bot, a bi-pedal humanoid robot capable of performing routine tasks, which would change the way people work around the world.

ServiceNow:

Hyperion believes that recent advancements in generative AI have unlocked another positive catalyst for the ServiceNow platform which should help the company to deliver better workflow results for its customers. The company has invested in AI talent over the past six years and is in the process of launching a number of generative AI features across a multitude of workflows. ServiceNow’s connector Integration Hub provides connectors to existing LLMs that companies can leverage to augment the ServiceNow platform. The company is also developing their own domain-specific LLMs (i.e. for IT Service Management and Customer Service Management, etc.) in partnership with Nvidia that will be trained on ServiceNow-specific data helping achieve greater accuracy and value beyond what general purpose LLMs can offer. At its Analyst Day held in May, the company showcased numerous use cases of their generative AI offering, successfully highlighting the meaningful productivity improvements achieved for end users. Increased productivity from generative AI offerings should allow ServiceNow to increase average selling prices of their installed base and be a net driver of revenue growth.

Microsoft:

Microsoft has been at the forefront of cutting-edge AI and ML, integrating these powerful, innovative technologies to improve product efficiencies and user experience. Microsoft’s AI platform, Azure AI, combines the fastest available graphic processing units with network architecture purpose built to enable AI model training and inference at scale. The company is also leveraging off its partnership with OpenAI, through the creation of next-generation AI models across consumer and enterprise products, including the Bing search engine, sales and marketing software, GitHub coding tools, Azure cloud and the Microsoft 365 productivity bundle. Microsoft remains committed to investing in the development of AI/ML technologies, announcing in January 2023 the third phase of the company’s long-term partnership with OpenAI through a multiyear, multibillion dollar investment, positioning the company to leverage off generative AI as an accelerant for transformation.

Intuit:

Data-driven innovation is core to Intuit’s success, and the company has recognised the importance of structured data capture for many years. Through its dominant competitive position across both its Small Business Group and Consumer Group, Intuit generates 400k customer and financial attributes per small business customer and 55k tax and financial attributes for each of its consumer customers. This large and unique data set is key to Intuit’s sustainable competitive advantage, and underpins its proprietary generative AI operating system (GenOS) and custom-trained financial LLMs. Intuit’s AI suite generate 58 billion ML predictions per day (730 million AI-driven customer interactions per year) and specialise in solving tax, accounting, marketing, cash flow, and personal finance challenges for its 100m+ customers around the world.

¹The Hyperion Global Growth Companies Fund (Managed Fund) changed its name from Hyperion Global Growth Companies Fund on 5 February 2021 in order to facilitate quotation of the fund on the ASX.

General Disclaimer:

This Report has been prepared by Hyperion Asset Management Limited (ABN 80 080 135 897 AFSL 238 380) (Hyperion) as the investment manager of the Hyperion Australian Growth Companies Fund (ARSN 089 548 443), the Hyperion Small Growth Companies Fund (ARSN 089 548 943) and the Hyperion Global Growth Companies Fund (Managed Fund) (ARSN 611 084 229) (each a Fund and together, the Funds). The Hyperion Global Growth Companies Fund changed its name to the Hyperion Global Growth Companies Fund (Managed Fund) on 5 February 2021 in order to facilitate quotation of the fund on the ASX. Interests in the Funds are issued by Pinnacle Fund Services Limited (ABN 29 082 494 362 AFSL 238 371) (Pinnacle) as responsible entity of the Funds. Pinnacle is not licensed to provide financial product advice.

Any opinions or forecasts reflect the judgment and assumptions of Hyperion and its representatives on the basis of information at the date of publication and may later change without notice. Any projections contained in this Report are estimates only and may not be realised in the future. Unauthorised use, copying, distribution, replication, posting, transmitting, publication, display, or reproduction in whole or in part of the information contained in this note is prohibited without obtaining prior written permission from Hyperion. Past performance is for illustrative purposes only and is not indicative of future performance.

The Product Disclosure Statement (PDS) and Target Market Determination (TMD) of each of the Funds is available at https://www.hyperion.com.au/. Any potential investor should consider the relevant PDS and TMD before deciding whether to acquire, or continue to hold units in, a Fund. For historic TMDs, please contact Hyperion Investor Services, Phone 1300 497 374 or Email investorservices@hyperion.com.au. The information in this Report has been prepared without taking account of any person’s objectives, financial situation or needs and is not intended as a recommendation or statement of opinion intended to influence a person or persons in making a decision in relation to investment. A financial adviser should be consulted before making any investment decision. This Report may contain the trade names or trademarks of various third parties, and if so, any such use is solely for illustrative purposes only. All product and company names are trademarks™ or registered® trademarks of their respective holders. Use of them does not imply any affiliation with, endorsement by, or association of any kind between them and Hyperion. Pinnacle and Hyperion believe the information contained in this Report is reliable, however no warranty is given as to its accuracy, reliability or completeness and persons relying on this information do so at their own risk. Subject to any liability which cannot be excluded under the relevant laws, Hyperion and Pinnacle disclaims all liability to any person relying on the information contained in this Report in respect of any loss or damage (including consequential loss or damage), however caused, which may be suffered or arise directly or indirectly in respect of such information. This disclaimer extends to any entity that may distribute this Report. No distribution of this Report will be made in any jurisdiction where such distribution is not authorised or is unlawful. Unless otherwise specified, all amounts are in AUD. Due to rounding, numbers presented throughout this note may not sum precisely to the total indicated and performance percentages may not precisely reflect the absolute returns.