AI Stocks: The ACTUAL Winners and Losers from the AI Revolution

In January, media firm Buzzfeed (NASDAQ:BZFD) made headlines after its CEO proposed using ChatGPT to help generate “cultural currency” and “inspired prompts” for his site. The following week, meme stock Helbiz (NASDAQ:HLBZ) would join the frenzy… without any detail on how an e-scooter business benefits from using chatbots.

Tongues would quickly start wagging.

“The AI stock mania is reminiscent of other speculative crazes in the world of tech,” wrote CNN’s Paul La Monica. “Remember when crypto-related stocks soared in 2021 and then tanked in 2022? And let’s not forget the epic rise of many dot-com companies in the late 1990s and their subsequent plunge in 2000.”

Yet, we know that artificial intelligence is here to stay; Mr. La Monica himself later conceded that AI “is no passing fad.” Much like the internet in the late-90s, we know that artificial intelligence creates tangible benefits… even if we’re not sure how to best use it yet.

That hasn’t stopped investors from unthinkingly rushing in. Shares of Buzzfeed would rise over 100% in the week after its ChatGPT announcement. And Helbiz would spike 250% for a brief period.

That means it’s more important than ever to separate the companies that claim they will benefit from AI, versus the ones that actually do. The internet of the 1990s couldn’t save outdated businesses from the dustbin of history. And artificial intelligence won’t either.

Chipmakers: Many Winners in the High-Tech World

The most obvious winners of the AI revolution are the chipmakers themselves. No matter which software firm eventually prevails, the top hardware makers must still play a role.

Winners: AI and Analog Chipmakers

The greatest beneficiaries of artificial intelligence are chipmakers that create GPUs (graphical processing units) and ASICs (application-specific integrated circuits) that help train AI models. Most hyped-up alternatives such as Tensor Processing Units (TPUs) and Data Processing Units (DPUs) are usually some combination of these two.

These types of chips can run multiple processes at once, making them far more powerful than traditional CPUs at training AI algorithms. GPUs are also becoming more scalable, thanks to a new generation of multi-node software.

As far as the GPU makers themselves, Nvidia (NASDAQ:NVDA) is the leader in the industry. The California-based firm holds a stunning 88% market share thanks to a massive R&D budget that can outspend all competitors. It’s a virtuous cycle that allows Nvidia to develop better hardware and more reliable drivers, which in turn sells more products. Analysts expect the GPU maker’s long-term growth rate to top 21%, according to Thomson Reuters.

However, buying top-dog Nvidia won’t come cheap. The company trades at 65X forward earnings, making it one of the most expensive hardware manufacturers in the world by that metric. Investors know that NVDA is the king, so much of the alpha has already been consumed.

Fortunately, investors have alternatives. First, the FPGA chip industry (field-programmable gate arrays) is a relatively profitable one that also benefits from AI. These are the low-power chips used to run AI models after they’re trained. Here, firms like Lattice Semiconductor (NASDAQ:LSCC) and Xilinx, a subsidiary of Advanced Micro Devices (NASDAQ:AMD), are profitable players that can be bought for lower prices than Nvidia. Lattice trades at 46X forward earnings, while AMD trades at 27X.

Next, investors looking further ahead should also consider analog chipmakers. These firms produce chips that are used in sensors and other IOT (internet of things) devices that connect computers with the outside world. Greater use of AI will naturally increase demand for these chips as applications from self-driving cars to smart homes become commonplace. The analog chip industry is also less competitive (and quite profitable) thanks to years of consolidation. Texas Instruments (NASDAQ:TXN) averages a 31% return on capital invested (ROIC), yet trades for only 21X forward earnings. Smaller competitor Analog Devices (NASDAQ:ADI) has averaged a 20% ROIC and trades at 18X.

Finally, specialized pick-and-shovel companies such as ASML (NASDAQ:ASML) also stand to gain. ASML, for instance, produces the machines that make AI chips, and can be bought at 33X forward earnings.

Losers: General-Purpose and Commodity Chipmakers

Meanwhile, general-purpose CPU companies will see a demand crunch, at least in relative terms. CPUs have more limited applications in artificial intelligence applications, and the industry’s cyclical nature makes it prone to post-overexpansion hangovers. Analysts estimate that Intel’s revenues will shrink almost 20% this year, and for 2026 profits to be less than half of 2019 levels.

Many commodity chipmakers are in even worse shape. Micron Technology (NASDAQ:MU), a firm that produces commoditized memory chips, has seen analysts cut long-term growth rates to -35.4%, according to data from Thomson Reuters. Though artificial intelligence promises to use more computing power, makers of lower-value chips will mostly see no benefit.

Software Companies: Mostly Good… If You’re High-Tech

It’s still too early to call winners with “generative” AI models. Chatbots — from OpenAI’s ChatGPT to Alphabet’s LaMDA — are relatively neck-and-neck in their problem-solving abilities. And the industry currently has surprisingly low barriers to entry; the internet is a trove of free training data, and talented engineers are quick to switch jobs. It’s a fool’s errand to guess whether Meta Platforms (NASDAQ:META) or Amazon’s (NASDAQ:AMZN) AI will come out ahead.

Nevertheless, winners are already emerging in other industries.

Winners: Data Analytics

The clearest corporate beneficiaries of AI are those that help users analyze, crunch and store data. Many of these companies — such as Salesforce (NASDAQ:CRM) and Workday (NASDAQ:WDAY) — have used existing products to establish a base, and then upsold customers on “value-added” AI products. Workday, for instance, has been able to help employers predict which workers are about to quit… since 2019. And Salesforce’s Einstein now provides salespeople with recommendations for the next best actions when making a sale. Rapid improvements in artificial intelligence will create even more “land and expand” abilities.

Other firms are more direct. Companies like cloud monitoring firm DataDog (NASDAQ:DDOG) and cybersecurity firm Okta (NASDAQ:OKTA) rely on machine learning to provide any product at all. Without it, their monitoring and authentication services will look much like the legacy systems they’re attempting to replace. And as artificial intelligence improves, products at these next-gen companies will stand out even more relative to legacy products. DataDog and Okta are expected to grow revenues by 32% and 41%, respectively, this year.

Losers: Business Process Outsourcing Firms

On the other hand, AI is squeezing lower-value software companies. And in particular, business process outsourcing (BPO) firms are feeling the heat. Former meme stock Exela Technologies (NASDAQ:XELA) now trades at a near-zero valuation. And Xerox’s BPO spinoff Conduent (NASDAQ:CNDT) has lost over 70% of its market value since spinning off in 2016. These companies specialize in offloading manual tasks to offshore workers, and AI tools are replacing everything from customer service chats to manual data entry.

Perhaps the only notable exception is Accenture (NYSE:ACN) — a consulting firm that’s long focused on higher-value BPO solutions. But even there, growth is slowing. Analysts expect Accenture’s earnings per share (EPS) growth to decline from the 16% range to single-digits this year.

Digital Media: A Mixed Bag

Winners: You and Me

In 2021, intense competition between Netflix (NASDAQ:NFLX) and other streaming services sparked “peak TV,” a period of incredible content creation. The number of original, scripted, adult-targeted programs jumped to 559, according to FX Research — straining the wallets of companies from Amazon to Disney (NYSE:DIS)

A similar gold rush is now happening in generative AI, particularly in digital art and chatbots. High-quality image generators from Midjourney to Stable Diffusion are available to anyone with an internet connection (and enough patience to wait in line). And Microsoft’s $10 billion investment in OpenAI and its ChatGPT is still (mostly) being given away for free to the public. (Google’s Bard might have to be given away, considering its issues).

The bonanza will last longer than most realize. The cost of running an AI model costs far less than… say… producing a hit TV show. And the competitive cloud industry means that cloud computing firms themselves are the ones using AI chatbots as loss leaders. Online storage might be a commodity. A high-quality chatbot running on it is not.

Losers: Content Creation Sites… And Many NFT Holders

Sadly, advancements in AI won’t help content creation companies — a competitive industry that has already exhausted the benefits of outsourcing. Websites like Forbes already routinely use freelancers in place of full-time staff. Only 10% of Buzzfeed’s staff belong to the NewsGuild labor union.

The inclusion of AI will only worsen the “race to the bottom” in digital media. Buzzfeed holds no monopoly over ChatGPT usage, so any financial benefit from using the service will be quickly erased by competitors doing the same. Only “AI-content natives” might benefit.

It’s the same story we saw in apparel in the 2000s with the rise of e-commerce: No clothing brand received an outright advantage, save for online-only companies.

Then there are NFT holders — investors who splashed out in 2021-2022 to acquire the rights to expensive digital art. Today, AI can create digital pieces by the thousands… and automatically mint them as NFTs if asked. It’s a comeuppance for those who once noted that “lots of NFTs are fairly dumb.”

Many younger investors might wonder what it was like to invest during the late-1990s tech bubble. With the introduction of AI chatbots and other generative models, these investors don’t have to dream any longer.

On the date of publication, Tom Yeung held LONG positions in GOOG and GOOGL. The opinions expressed in this article are those of the writer, subject to the InvestorPlace.com Publishing Guidelines.