C3.ai (AI 0.08%) went public at $42 on Dec. 9, 2020, started trading at $100, and soared to an all-time high of $177.47 just two weeks later. Investors were initially impressed by its enterprise AI algorithms -- which could be plugged into a large organization's existing software to automate and accelerate certain tasks -- and its rapid growth rates. Its ticker symbol also contributed to its popularity as a meme stock.

But as of Monday, C3.ai's stock was trading below its IPO price in the $38 to $40 range. The bulls retreated as its growth cooled off, it racked up steep losses, and rising interest rates popped its bubbly valuations. After growing its revenue at a compound annual rate of 40% from its fiscal 2019 to its fiscal 2022, its revenue only rose by 6% in its fiscal 2023 (which ended April 30, 2023) as macro headwinds and pessimism drove many companies to rein in their spending.

Image source: Getty Images.

In its fiscal 2024, C3.ai's revenue rose by 16% as more organizations installed its AI algorithms in their efforts to hop aboard the AI bandwagon, and management expects an acceleration to between 19% and 27% growth in fiscal 2025 as that market expands. Analysts expect its revenue to grow at a compound annual rate of 20% from fiscal 2024 to fiscal 2027.

That sounds promising, but there are three bright red flags for investors.

First, C3.ai generates more than 30% of its revenue through a joint venture with Baker Hughes (NASDAQ: BKR). That deal will expire by the end of fiscal 2025, and it hasn't been renewed yet. Second, it expects to remain unprofitable as it develops more generative AI tools. Lastly, C3.ai has gone through three CFOs since its IPO. As it rotated through those executives, it repeatedly changed its formula for counting its customers and cannibalized its own subscription services with consumption-based plans.

Given these challenges, C3.ai's stock doesn't seem like a bargain at 10 times this year's sales. So instead of investing in that wobbly business, investors should consider buying two cheaper AI-driven stocks: Micron (MU -3.54%) and IBM (IBM -0.83%).

1. Micron

Micron manufactures DRAM and NAND memory chips. It's not the biggest player in either market, but it produces denser chips than most of its larger competitors. Those memory chips can't process AI tasks on their own, but the memory they provide is essential for supporting demanding AI applications. Higher-density DRAM chips allow data centers to store more short-term data, while higher-capacity NAND chips are used to store more long-term data in solid-state drives (SSDs).

Micron's business generally ebbs and flows with the cyclical memory market. In its fiscal 2023 (which ended in August 2023), its revenue plunged by 49% and it posted an adjusted net loss. That slowdown was caused by the cooling PC market, the end of the 5G upgrade cycle in the smartphone market, and the prioritization of AI-oriented GPUs over other types of chips in the data center market.

But in fiscal 2024, Micron's revenue surged by 62% as it turned profitable again. That recovery was supported by the stabilizing PC and smartphone markets, as well as data center operators purchasing more high-capacity SSDs and high-bandwidth memory (HBM) chips to support AI applications.

In fiscal 2025, analysts expect Micron's revenue and adjusted EPS to grow by 52% and 587%, respectively, as that cycle continues. Those would be incredible growth rates for a stock that trades at 12 times forward earnings. This growth phase won't last forever, but the stock could still have a lot of upside potential.

2. IBM

IBM is often viewed as a slow-growth tech giant rather than an exciting AI play. From 2012 to 2020, its annual revenues declined from $107 billion to $55 billion as it divested itself of several of its major businesses, struggled with sluggish growth in its legacy enterprise hardware, software, and IT services divisions; and failed to keep pace with its industry peers in the expanding cloud infrastructure and services markets.

But from 2020 to 2023, IBM's revenue and EPS grew at compound annual rates of 4% and 9%, respectively. That recovery was driven by two main tailwinds. First, it hived off its slower-growth managed IT infrastructure services unit into a new company, Kyndryl (NYSE: KD).

Second, it focused on expanding the presence of Red Hat -- its open-source software subsidiary -- in the hybrid cloud and AI markets. Instead of going head-to-head against cloud giants like Amazon and Microsoft, IBM developed more open-source AI tools for processing the data that flowed between public cloud platforms and on-site private clouds. Those "hybrid" cloud deployments were popular with large companies that weren't ready or willing to migrate all of their data from local servers to cloud-based platforms.

From 2023 to 2026, analysts expect IBM's revenue and EPS to grow at compound annual rates of 4% and 5%, respectively. Its stock is reasonably valued at 21 times forward earnings, its dividend offers a decent yield of 3% at the current share price, and it could outperform flashier AI stocks like C3.ai over the long run.