What is artificial intelligence?
If you ask IBM (NYSE:IBM), “at its simplest form, artificial intelligence is a field, which combines computer science and robust datasets, to enable problem-solving” while also encompassing “sub-fields of machine learning and deep learning, which are frequently mentioned in conjunction with artificial intelligence”.
IBM’s opinion on the subject matters- despite not being at the vanguard of AI innovation today, Big Blue was an early authority on the subject, having been consulted on Stanley Kubrick’s big-screen adaptation of Arthur C. Clarke’s classic 1968 novel 2001: A Space Odyssey.
Keen observers will note that the central antagonist of Clarke’s parable, the homicidal computer HAL (short for Heuristically programmed ALgorithmic computer) has a name comprising mere one-letter shifts from IBM.
It was chalked up as a coincidence, but this piece of trivia goes to show that the concepts of AI machine learning have a longer history than one might initially think.
In fact, Alan Turing’s seminal research paper Computing Machinery and Intelligence posed what turned out to be an influential question: “Can Machines Think?”
Science, pop culture and commerce have dedicated infinite work hours to answering this question, out of which some of the largest megacorporations of all time have been born.
Today, global tech titans like Microsoft Corporation (NASDAQ:MSFT), Google parent Alphabet (NASDAQ:GOOGL) Inc (NASDAQ:GOOG), Meta Platforms Inc (NASDAQ:FB) and Nvidia Corporation are all seeking to claim their stake in the current AI boom.
This AI arms race is leading to stratospheric levels of capital expenditure in the hopes of even greater shareholder returns in the long run.
But where do the household-name Big Tech players currently stand in the commercial AI sector and how much have they spent to get there?
Let’s have a look.
Microsoft
Microsoft’s early investment into OpenAI, the developer of ChatGPT, the large-language model effectively synonymous with the 2023 AI boom, could turn out to be one of the wisest, or luckiest, in history.
Microsoft's investment in OpenAI began in 2019, when the tech giant invested $1 billion in the AI research company. In January 2023, Microsoft announced another multi-year, multi-billion dollar investment in OpenAI, following rumours of a $10 billion deal.
Today, Microsoft owns 49% of the company.
The two companies have a close partnership, with Microsoft providing OpenAI with access to its Azure cloud computing platform and other resources. Microsoft has also incorporated OpenAI's AI models, such as GitHub Copilot, DALL·E 2, and ChatGPT into its own products and services.
Microsoft's investment in OpenAI is seen as a major bet on the future of artificial intelligence and represents something of a revolution for the legendary Redmond computer company.
OpenAI is reportedly in discussions to sell shares held by current employees at a staggering valuation of US$86 billion, or which Microsoft’s stake would be worth $43 billion.
It already appears to be paying off for Microsoft, with revenue up 13% year-over-year to $56.5 billion in the latest quarter, driven by its cloud division and its adoption of AI services powered by OpenAI.
Facebook (NASDAQ:META) parent Meta Platforms Inc has spunked over $30 billion – and counting – on developing Mark Zuckerberg’s vision of the metaverse.
This astronomical spending spree kicked off in late-2021, when many were convinced that digital twins and 3D sandbox playgrounds were the next big thing.
Meta’s Horizon Worlds famously launched to universal ridicule, though recent tech showcases suggest that we were a bit too early to laugh.
Regardless, there’s a case to be made that Zuckerberg put all his eggs in the metaverse basket, leaving him with his pants down when sentiment soured on the metaverse in favour of shiny new AI technology.
This led to a defensive second-quarter earnings call in July, where he attempted to espouse Meta’s AI credentials.
“I've said on a number of these calls that the two technological waves that we're riding are AI in the near term and the metaverse over the longer term,” said Zuck, hoping to convince shareholders that he was on top of, rather than simply following, the trends.
In Zuck’s defence, Meta has developed its own ChatGPT killer in the form of Llama.
In July 2023, Meta announced the release of Llama 2, a successor to the original Llama model, a larger and more advanced model than Llama available freely for research and commercial use.
But monetisation does not seem to be part of Meta’s AI roadmap, at least for now. Rather, it is likely to be a considerable expense.
“The other major budget point that we're working through is what the right level of AI capex is to support our roadmap,” said Zuck in July. “Since we don't know how quickly our new AI products will grow, we may not have a clear handle on this until later in the year.
“Investments that we've made over the years in AI, including the billions of dollars we've spent on AI infrastructure, are clearly paying off.”
For Meta, AI currently plays more of an internal role in advertising and engagement across its suite of social media platforms.
Apple
The world’s largest company, tech or otherwise, is not an AI powerhouse, and has yet to show its hand in the sector.
A case could be made that Apple Inc (NASDAQ:AAPL). suffered a bout of Meta syndrome, putting its chips on the metaverse by diverting development resources to the Apple Vision Pro headset.
It’s not yet clear what the appetite for a $3,500 “spatial computer” with less battery life than an average feature film is, though esteemed Apple correspondent for Bloomberg Mark Gurman recently wrote on Apple’s scramble to jump aboard the AI train.[
“I can tell you in no uncertain terms that Apple executives were caught off guard by the industry’s sudden AI fever and have been scrambling since late last year to make up for lost time,” he wrote, while an insider stated: “There’s a lot of anxiety about this and it’s considered a pretty big miss internally.”
Internally, Apple is building an LLM called Ajax and has an early-stage chatbot with the catchy name ‘Apple GPT’ in the works.
Technically a consumer hardware play, Apple is under pressure from a global downturn in smartphone sales.
The group is expected to report a year-over-year decline in sales but higher earnings for the fourth quarter when it releases its latest financial results after the market close on Thursday, November 2.
Apple’s next line of Mac is also due to be unveiled during an October 30 launch event. Will Tim Cook shoehorn a surprise AI announcement in? Maybe!
Cook, by the way, is IBM’s most prestigious alumni, though evidently showed little interest in carrying the AI flame first lit by his former employer.
Nvidia
Despite being one of the largest companies in the world, Nvidia Corporation’s 30-year history as a graphics processing giant largely went under the radar, at least when compared to its Big Tech contemporaries.
This all changed in 2023, when Nvidia emerged as the number-one AI play on the capital markets, with its valuation shooting well above the one-trillion-dollar mark.
The story of Nvidia’s cementation into the AI industry is a long one, but the inflection point came when boffins realised that its parallel graphics processing units were perfectly optimised to handle AI applications.
Cue OpenAI using 10,000 of these GPUs to train its landmark ChatGPT LLM, and suddenly Nvidia’s hardware was the hottest commodity in the tech world.
So much so that they became subject to US export controls limiting their distribution in China for fears that they would advance the communist state’s military-industrial complex to dangerous levels.
Nvidia is in an enviable position. Not only is its hardware powering the AI revolution, but its software stack is becoming increasingly integral to the industry too.
Competition is also scarce. While Microsoft, Meta and Alphabet attempt to outclass each other at the consumer level, Nvidia is there as the pick and shovel, providing the framework for all of its contemporaries to build upon.
Nvidia’s dominant position in the AI space was keenly reflected in its stellar earnings call in August, which marked the first time in history that a semiconductor firm guided revenues $4 billion above consensus.
Given Nvidia's pivotal role in AI, one market analyst contended that the company could potentially see demand in the $30 billion per quarter range.
It will take a big shift in the market for Nvidia to be knocked off its perch.
To the end, Nvidia’s closest GPU competitor Advanced Micro Devices, Inc (NASDAQ:AMD) recently acquired artificial intelligence startup Nod.ai.
This was read as a preliminary step in competing with Nvidia in the burgeoning AI field, but it is still early stages, and Nvidia’s early-starter advantage cannot be understated.
Google provided the first true comedy of errors when parent company Alphabet Inc announced its ChatGPT killer in February. That made its name Bard rather fitting.
Bard is built on Google's LaMDA (Language Model for Dialogue Applications) family of large language models.
LaMDA was first announced in 2021, but it was not released to the public until Bard was released in 2023.
The launch did not go amazingly, having answered a number of questions with factual inaccuracies.
Bard appeared to have been rushed out the gates in order to capitalise on ChatGPT’s successes.
More recently, there are reports that Google insiders themselves are questioning Bard’s utility.
Earlier this month, senior product manager for Bard Dominik Rabiej was quoted on a private Google Discord channel stating: “My rule of thumb is not to trust LLM output unless I can independently verify it”.
Rabiej suggested limiting people’s use of Bard to “creative / brainstorming applications”, though was also useful for coding, said Rabiej, “since you inevitably verify if the code works”.
Not a great look for a supposed ChatGPT killer.
Google also underwhelmed in its latest earnings call, though this was put down to the underperformance of its cloud computing enterprise, which has always trailed Microsoft’s and Amazon’s services.
Like its contemporaries, Google does not disclose how much it is spending on internal and public-facing AI projects, In 2023, though parent company Alphabet did increase its capital expenditure budget by $13 billion in 2023, with a significant portion of that investment going undoubtedly going towards AI initiatives.
It is also worth noting that Alphabet bought Nest Labs for $3.2 billion in 2014 to build out its cutting-edge line of smart home products.
Amazon (NASDAQ:AMZN)
Like Apple, e-commerce giant Amazon is not exactly synonymous with AI innovation.
AI was only mentioned three times in Amazon’s last annual report, with chief executive officer Andy Jassy stating that LLMs and generative AI “are going to be a big deal for customers, our shareholders, and Amazon”.
However, though Amazon is less visible in the AI space, its market-leading (some regulators might say anti-competitive) cloud computing service Amazon Web Services (AWS) has successfully monetised the Amazon Lex conversational AI bot since 2017.
Like AWS’s core hosting service, Lex is an economy of scale, charging $0.004 per speech API request and $0.00075 per text API request.
Delta Airlines (NYSE:DAL), Bank of America (NYSE:BAC), Hilton Hotels and other major companies use this service for their customer service chatbot interactions.
This is not exactly tapping into the cutting-edge AI and machine learning zeitgeist that we’re talking about here, and anyone forced to use such chatbots might say they’re doing harm than good.
Amazon remains the premier cloud computing play, but AI? Not so much. However, there is no doubt that AI is playing an increasingly important role internally at Amazon, as it is at Netflix (NASDAQ:NFLX), SAP, Salesforce and every other major tech group.
Such is the reality of AI: Though not every tech company is an AI play, every tech company needs AI to thrive.