AI Era: Hardware frenzy, software loneliness

Under the wave of AI, hardware manufacturers like NVIDIA and Dell are making a fortune. However, application software companies that once shone in the era of mobile internet now appear quite desolate, unable to attract capital investment.

In the first quarter of this year, Dell's AI server revenue hit a new high. Morgan Stanley stated that Dell has excellent engineering capabilities in the field of AI servers, and due to the technical complexity of this field, Dell is one of the few winners among server manufacturers. So far this year, Dell's stock price has risen by nearly 100%.

As for NVIDIA, it goes without saying. This Tuesday, NVIDIA surpassed Microsoft and Apple to become the world's most valuable company by market capitalization. The company's stock price has risen by over 170% so far this year.

In contrast, software application companies seem to be stagnant in the AI era. Although application software manufacturers are very keen on building AI tools, it is not clear whether anyone is willing to pay for them.

For example, AI tools from software service providers can help you write documents in Salesforce, summarize employee skills in Workday, or create images in Adobe's Photoshop based on prompts... However, Bloomberg pointed out that in this year and a half of AI frenzy, these tools have not brought much revenue to these software service providers. For large application software companies, AI-related sales will not appear on the profit and loss statement until next year or the year after.

In many cases, software companies cannot even decide how to charge. Some companies use AI as a selling point for high-priced subscriptions, while others, such as remote meeting software provider Zoom Video Communications, simply bundle AI features for free.

Investors have lost patience. The iShares Expanded Tech-Software Sector ETF is a common indicator for measuring this industry, and after rising by 59% in 2023, the ETF has only risen by about 3.5% this year.

In the era of mobile internet, applications dominated the market, with computing power and infrastructure accounting for a smaller proportion. Because the power of the internet lies in "connection," value needs to be realized through applications. However, in the AI era, the underlying aspects - computing power and infrastructure - account for a higher proportion because it focuses on "generation."

KeyBanc analyst Jackson Ader stated that most of the spending in the AI industry is currently flowing towards hardware or cloud infrastructure used for training and deploying large models. Ader also pointed out that in many cases, this spending comes at the expense of traditional software suppliers. He said that most Chief Information Officers are adopting a "wait-and-see" attitude towards AI-based application software.

Furthermore, discussions about "AI replacing programmers," "AI replacing drawing tools," "AI replacing customer service," etc., are ongoing... A long-term concern is that the anxiety of unemployment in the software industry will spread A research report released by Citigroup this week also pointed out that about 36% of positions in the software platform industry have a high potential to be replaced by "AI".

Goldman Sachs analyst Kash Rangan raised this question to Adobe management during last week's earnings conference call. He asked whether generative AI is good enough to become the terminator of the creative process. Therefore, (due to AI replacement), we no longer need creative individuals. Rangan pointed out that this is a major issue of debate among investors