Introducing GXAI: Capture the AI Revolution
Achieving and using Artificial Intelligence (AI) has been one of the longest-running pursuits in computer science. With the release of ChatGPT, the industry hit an inflection point by which the general public became aware of AI’s far-reaching possibilities. Since then, we have seen practically all of the world’s most influential companies invest record amounts of capital toward AI implementation.1 As such, it appears that AI has finally become ‘mainstream’ and has undoubtedly become one of the most disruptive themes in the world – a megatrend that investors cannot ignore. The Global X Artificial Intelligence ETF (ASX: GXAI) aims to capture this megatrend by tracking the performance of global companies best positioned to capitalise on its growth and proliferation.
What is AI Today?
AI is the concept of machines performing tasks that once required human intellect to complete, and as such, the range of programmable tasks that constitute AI is massive. In order to break down such a diverse ecosystem of uses, each AI system can be segmented into one of three ‘levels’:
- Level 1: AI is a catch-all term which describes any system that can perform self-learning and use the learnt knowledge to produce an outcome.
- Level 2: Machine Learning is a more advanced form of AI which allows the machine to access structured data, apply algorithms to derive valuable insights, and apply what they learned to other scenarios or new data sets.
- Level 3: Deep Learning is currently the most advanced form of AI. Deep learning attempts to emulate the human brain and uses artificial neural networks to make sense of unstructured data. This allows the AI to form its own ‘understanding’ of events and generate outputs outside of its original human programming.
A common theme across all three levels is that the power of AI comes from an availability in data. As digital data continues to accumulate, deep-learning models, with their ability to understand real-world, unstructured data, will likely grow at unprecedented rates. It is these models that currently represent the bulk of opportunities for AI across the economy.
How Will AI Impact the Economy?
AI’s broad applicability represents a platform shift in the making, one that will have a broad influence on the adoption of technology across the economy. According to one report, AI could contribute up to US$16 trillion to global GDP in 2030 with roughly US$9 trillion originating from consumption side-effects and the other US$7 trillion coming from increased productivity and efficiency.2 For context, that would equate to almost 14% of global GDP, more than the combined growth of China and India today.3
At a fundamental level, the investment opportunity for AI is the disruption of existing business models. Roughly 56% of companies utilise AI in their operations today, but most are only engaging in limited implementation.4 Despite this, we have seen numerous large-scale industry transformations. Below are a few examples:
- Cybersecurity has seen a massive efficiency boost with the introduction of AI machine learning. Today, AI’s ability to parse large datasets and generate actionable insights are crucial to the industry. Crowdstrike’s ‘Threat Graph’, for example, can handle more than a trillion cyber events and analyse over 15 petabytes (15 million gigabytes) of data per day. Zscaler, a cloud-based cybersecurity platform, also performs 250,000 security updates daily – with its AI responsible for ensuring secure customer access to everything from corporate emails to web applications.
- E-commerce and the digital economy will likely see large cost reductions and improvements in efficiency as natural language processing tools such as chatbots and speech recognition replace the need for human-to-human interfaces and provide live sales assistance. Last year, furniture retailer Temple & Webster announced that all pre-sales product queries would be powered by ChatGPT after the company saw a 50% improvement in market awareness and up to 350% improvement in sales targeting metrics.5
- Agricultural industries have been surprisingly fast adopters of AI software. Tractor company, John Deere, revealed its first fully autonomous electric tractor for large-scale production in 2022 – the vehicle operates through deep learning on satellite imagery and can even react within 100 milliseconds when detecting objects in its path.6 Also emerging is ‘precision agriculture’, whereby AI networks monitor every detail from soil quality to incoming weather patterns and even weed growth. Carbon Robotics, for example, is a Detroit based start-up which has developed farming bots that pass over farmland, identify weeds through AI recognition, and then destroy the emerging growth with high-powered lasers.7
The above-provided examples display the extent to which AI has already embedded itself into multiple industries despite only scratching the surface of its immense potential. As AI advancements create further use cases, AI demand will grow rapidly as a result.
Who Will be the Winners of AI?
While most global enterprises are rushing to extract value from AI implementation, we believe it is actually those that provide the means of accessing AI who will be the biggest winners of the megatrend. As the saying goes, ‘During a gold rush, sell shovels’. Global X has identified three groups of ‘shovel sellers’ which we believe will likely benefit the most from the impending transformation. All listed companies mentioned below feature in the Global X Artificial Intelligence ETF (ASX: GXAI). See GXAI’s full holdings here.
Unsurprisingly, tech giants rank well among these categories, with most of the magnificent seven participating in at least two of the three. Being leaders of today’s technology frontier, they are well equipped with talent, existing infrastructure, and excess capital for required R&D. Furthermore, most big tech firms are in possession of vast pools of proprietary data – a key resource for AI development.
More broadly though, we believe AI developers will be the major beneficiaries of AI advancement – judging by investments, big tech agrees. AI startup investments have grown rapidly over the past decade, but the market truly hit its stride last year when it received almost US$50 billion in cash injections, marking one of the biggest investment cycles in recent history.8,9 Big tech contributed to more than half of that total investment – Microsoft was the biggest investor of the year, piling US$13 billion into its widely-reported partnership with OpenAI; Google and Amazon were also notable big hitters, investing almost US$20 billion into AI Firm ‘Anthropic’, as well as OpenAI over 2022-23.
Outside of VC-style acquisitions and investments, big tech firms have also spent significant capital toward internal AI capabilities, positioning themselves as AI developers for the future. Last year the Mag-7 spent more than US$160 billion in investments despite a hostile interest rate environment, and the year before (when ChatGPT was released) the group totalled US$170 billion.10 Reports also forecast that investments in generative AI alone will account for 12% of global total tech expenditure by 2032.11 With such huge commitments toward the industry, it is apparent that the tech titans of the world are all in recognition of AI’s huge economic potential.
Last year saw cloud computing companies overwhelmingly shift their investment objectives away from general computing to AI computing. Enterprise datacentre spending grew almost 27% last year and will reach US$286 billion in total for 2023.12 Looking into the future, that figure is expected to grow at an annual rate of almost 25% and reach close to US$600 billion by 2027.13 Hyperscalers such as AWS, Google Cloud, and Azure, will play a large role in that capex, with one report suggesting that the group could represent nearly half of the world’s global data infrastructure spending over the next few years.14 Similar to how the internet saw the initial rise of datacentres, the proliferation of AI will open a new age of growth for cloud service providers as they lower AI’s cost of entry and become the gateway to AI implementation.
Past performance is not a reliable indicator of future performance. Forecasts are not guaranteed, and undue reliance should not be placed on them.
AI hardware firms are fundamental to the megatrend. Generative AI and deep learning models use massive amounts of compute in training and operations, which is powered by AI chips provided by firms such as Nvidia, Intel, Broadcom, etc. As AI grows in prominence, AI chipsets will play a much larger role in datacentres. Datacentres have historically spent around US$220 billion in updated hardware annually, and thanks to the cyclical nature of datacentre systems (equipment is generally replaced every four years) that means roughly a trillion dollars’ worth of computational infrastructure will be up for replacement, a generational opportunity for AI hardware to gain ground in the industry. Global X expects demand for AI chips to approach US$165 billion by 2030, representing a CAGR of more than 30% over the next 10 years.15
Past performance is not a reliable indicator of future performance. Forecasts are not guaranteed, and undue reliance should not be placed on them.
Why Invest in AI with ETFs?
Stock selection is infamously difficult, especially when it comes to picking the winners of a new, emerging industry. With each thematic trend, many companies rise to become industry titans, but just as many fall to obscurity and fail to provide meaningful returns.
Thematic ETFs provide a more holistic approach for investing in emerging technologies by capturing the full value chain. For investors who wish to capitalise on the AI megatrend but are uncertain which stocks may emerge as winners – Global X Artificial Intelligence ETF (ASX: GXAI) may be the solution. GXAI seeks to provide exposure to the full investible AI universe, from small-cap semiconductor firms to mega-cap tech giants.
Related Funds
GXAI: The Global X Artificial Intelligence ETF (ASX: GXAI) invests in global companies involved in AI development, AI-as-a-service, provide AI compute power, or design and manufacture AI hardware.