Yahoo Finance AI’s path in 2024: Another year of growth instead of monetization
Artificial intelligence has captured the world’s imagination, but for 2024, investors shouldn’t expect to see much in the way of financial tailwinds yet — outside a couple of exceptions to the rule.
Direct monetization of AI services will remain out of reach for most companies, Gil Luria, managing director at D.A. Davidson, told Yahoo Finance. While businesses are ramping up their AI usage, software makers will likely include AI features as a perk of their existing offerings or as a way to prompt upgrades instead of a paid standalone product.
“The tools built around generative AI are compelling, but they are in their infancy. They're still working out the kinks,” said Luria. “The inclination from companies … will be to bundle these offerings in order to drive adoption, so we're still probably a little bit farther away from being a material growth driver.”
Tell that to investors who have become enthralled with the AI trade.
Software ETF IGV is up roughly 50% over the past year, compared to a 34% gain for the Nasdaq Composite (^IXIC) and 20% for the S&P 500 (^GSPC). The outsized performance is mostly driven by AI excitement, said Luria. "We believe most software companies will face a headwind to their stocks if they don't deliver tangible AI results [this year]."
Microsoft still leads the industry, Salesforce well positioned
The outliers are Microsoft (MSFT) and Adobe (ADBE) — both D.A. Davidson’s top picks for mega caps due to their advanced AI abilities and solid businesses outside of the field.
Adobe, a leader in text-to-image generation, has incorporated generative AI features across its creative suites, with a generative AI credits system that entices users to pay more to use those functions.
Microsoft, with its Copilot services, is charging users extra for accessing the AI assistants embedded within Microsoft 365, GitHub, and more.
But the real AI cash cow will remain its cloud service, Azure, which provides access to OpenAI's and others' large language models (LLMs) that developers can use to build their own applications.
While monetization might not come right away, software that aids client-facing roles is well positioned for the future. Generative AI’s potential has been touted as almost limitless, but the proven use cases currently focus on text generation, text summation, and text-to-image generation, per Luria.
These use cases are most relevant in tasks like pre-populating emails, summarizing calls, and answering common questions — all functions built into Salesforce’s (CRM) AI assistant, Einstein Copilot, which is targeting the end of February for a public launch.
Morgan Stanley recently upgraded the company's stock to a top pick, citing its consistent revenue growth and relatively muted trading multiples compared to its peers.
Data is worth gold in the AI age, so investors should look for firms that lead the field in data access and integration. Salesforce, for example, already hosts a wealth of data from companies using its products to automate customer service, marketing, and other tasks.
In December, the enterprise SaaS giant announced a new capability to import unstructured data — think emails and phone calls instead of entries in databases — into its Data Cloud product, which connects data from its various software offerings.
More data is the solution to hallucinations, or false answers.
“One of the most common ways to improve a hallucination or prevent a hallucination is to provide more context,” Patrick Stokes, executive vice president of product and industries marketing at Salesforce, told Yahoo Finance. “If we can connect the business context into the prompt, the LLM now has the additional information it needs.”
The future: Smaller models, bots talking to bots
Though large, universal knowledge models like ChatGPT grab all the attention, companies pioneering more efficient models will be the ones raking in the cash. Workers at lamp makers don’t need help drafting emails about the Roman Empire. Smaller models containing only the pertinent information would cost less to operate — and thus, easier for businesses to profit from. Such efforts are underway at Microsoft and Salesforce, among others.
“We learned that the company has recently formed an internal team to develop smaller AI models (SLMs) to make AI cheaper to run and seemingly reduce dependencies on OpenAI. These SLMs, such as Phi-2 and Kosmos-2 are intended to rival performance of larger models with significantly fewer parameters,” Citi analysts wrote after attending Microsoft’s AI tour last week.
What’s in enterprise AI’s future? One day soon, we may see bots talking to bots — an AI assistant managing your calendar collaborating with an AI assistant handling your client information — as platforms become more interoperable.
“It's like the assistants start helping each other to help the end user,” predicted Nick Johnson, Salesforce’s senior vice president of strategic partnerships.
If these assistants can program profits for their masters, investors would be all the better.
Jennifer Wang is a senior editor at Yahoo Finance.
Click here for the latest technology news that will impact the stock market.
Read the latest financial and business news from Yahoo Finance
