Meta announced this year that it would invest $35 billion in artificial intelligence (AI), signaling an aggressive approach to the escalating technology arms race.
This huge investment raises critical questions about the future of AI development and its economic viability. Industry experts are currently debating the scope and impact of this funding, when these investments will generate a return on investment (ROI), and how they could reshape Big Tech's revenue model. is investigating. With potential strategies ranging from subscriptions to advertising, Meta's move could set a new precedent for how tech giants leverage advances in AI.
“Right now, there doesn't seem to be an end to the arms race,” Mudhu Sudhakar, CEO of generative AI company Icela, told PYMNTS. “AI is clearly a key strategic focus. Think of it like moving from on-premises to the cloud or from desktop to mobile. It's a trend. So for mega-tech companies like Microsoft, Google, Meta, and Amazon, missing out on AI would be disastrous. This is why they're spending billions of dollars in capital investment.”
Growing AI investment
In Meta's recent earnings report, the company increased its spending by $5 billion to develop new AI products for consumers, developers, enterprises, and hardware manufacturers, instead of increasing the return on investment from AI. It became clear that there was.
Investments in the company's AI and Metaverse development arm, Reality Labs, are expected to reach $35 billion to $40 billion by the end of the year.
CEO Mark Zuckerberg also spoke last week about the launch of Meta AI, the latest version of Meta's AI assistant. This AI is powered by Meta Llama 3, the latest update to the company's large-scale language models.
Even as Meta and other companies invest heavily in AI, observers say a return on investment may be remote.
“We are in the investment phase of the AI cycle,” Sudhakar said. “Therefore, it is unrealistic to expect widespread ROI. Another historical example is the early days of the Internet. Experimentation with use cases, education, adoption, and large investments in infrastructure is required. The same thing is happening with AI, but there's one area where there's a clear ROI: infrastructure providers, especially Nvidia. But over time, the ROI expands to things like applications. .”
Costs can hinder profits
While excitement and investment in artificial intelligence is growing, Sudhakar pointed out that this technology, especially generative AI, is very expensive.
“There are costs for GPUs, data science talent, data center expansion, energy usage, data management, etc.,” he said. “But the good news is that costs are starting to come down and that will help monetize. For example, this week Snowflake announced its own LLM, which costs just $2 million to train and only requires GPUs. It was 1,000 pieces.
The value of AI comes from internal efficiencies and cost savings, not just the products and services sold to end users, Lars Nijman, chief marketing officer at CUDO Compute, told PYMNTS.
“Certainly, some applications, such as fraud detection, chatbots, and better ad serving algorithms, quickly deliver clear and direct financial benefits,” he said. “However, for basic research projects, ROI can be more nebulous, measured in long-term innovation benefits and measured over years rather than quarters.”
Sudhakar said Big Tech companies are likely to adopt mixed business models to profit from AI models. Initially, companies with sufficient cloud infrastructure can quickly benefit by monetizing their investment through a consumption-based model for cloud services. This often involves accessing large-scale language models (LLMs) and small-scale language models (SLMs) through APIs. Additionally, there are opportunities to cross-sell development tools and various other services within this ecosystem.
Additionally, as Sudhakar pointed out, AI applications could be sold as subscriptions. For example, OpenAI has had great success with its ChatGPT system, while Microsoft has had similar success with its GitHub Copilot and is exploring the potential of its Office 365 Copilot product. Beyond subscriptions, advertising has become a viable revenue stream, especially for companies like Google and Meta, who can leverage extensive ad infrastructure and AI enhancements to optimize ad revenue.
“But Google is in a tough position,” he says. The company's core search business relies heavily on compensation for links. However, if a generative AI app can provide more information in his one response, this revenue stream will come under pressure. ”
Yigit Ihlamur, general partner at Vela Partners, told PYMNTS that revenue will skyrocket once advertising is powered by AI.
“For example, when you use ChatGPT to book a flight, you’ll see organic and sponsored search results,” he said. “We imagine technology companies will further monetize their models through paid subscriptions, paid developer-centric APIs, and marketplace transactions.”
