Over the past week, rumors have been circulating on the internet that may have caused some artificial-intelligence (AI) winners like Nvidia (NVDA 6.18%) to sell off.

Last week, the Taiwan Business Times cited "rumors in the market" that Microsoft (MSFT 1.82%) CoPilot subscriptions were not as strong as originally forecast, and that ChatGPT usage was down over the summer. That may have led Microsoft to revise down its orders for Nvidia H100 GPUs, which are the main chip used for AI training.

That certainly wouldn't be good, as AI stocks like Nvidia and Microsoft have taken off in 2023 on the prospect for a multiyear high-growth period.

However, I think it seems more likely than not these rumors are just that: rumors. Not only have several other recent data points contradicted the rumor of an AI overbuild, but at least one Wall Street analyst thinks the recent rumor-inspired dip in these stocks is a big buying opportunity.

Analyst channel-checks are actually growing more optimistic

On Monday, Morgan Stanley analyst Joseph Moore wrote a note calling the pullback in Nvidia shares a "buying opportunity." In contrast to the rumors of order cuts, Moore said his firm's recent channel checks showed demand for Nvidia H100s were still "well above" current supply, even as supply has improved.

This was true across the general supply chain and multiple customers, not just Microsoft. Microsoft is just one customer, and its OpenAI subsidiary grabbed the early lead in the AI races. However, the popularity of ChatGPT is spurring a rush of investment from AI startups and other large cloud companies alike. So, even if the Microsoft rumors were true, which they may not be, a slight pullback in orders after a massive jump would not necessarily portend a big downturn for Nvidia, which serves many others.

Yet even Microsoft demand seems solid. s Moore wrote, "Supply chain reports have Microsoft pushing for more product than they are currently getting, at least tactically, so we are very comfortable there isn't a near term air pocket with that customer."

Recent industry updates suggest no slowing in AI investment

Besides Moore's note, other data points also suggest stronger, rather than weaker, demand for AI investment generally, which will benefit Nvidia.

On Monday, OpenAI announced a new update for ChatGPT touting improved voice and image recognition. Customers can now have a voice-based rather than text-based conversation with ChatGPT thanks to new voice recognition capabilities, or customers can now show an image to ChatGPT to help frame what you're asking. For instance, you can now show ChatGPT a picture of an appliance that's not working, or snap a photo of what's in your fridge and have ChatGPT suggest a recipe.

It all sounds like the H.A.L. computer from the movie 2001: A Space Odyssey. That's fascinating, and also kind of scary, given what happens in that film. But regardless, these voice and image recognition updates, powered by ChatGPT 3.5 and ChatGPT 4.0, will also require lots and lots of Nvidia chips to run.

In addition to Microsoft and OpenAI's ChatGPT, Alphabet (GOOG 9.96%) (GOOGL 10.22%) is also working on its own large multimodal model called Gemini, which should be out soon. And also on Monday, Amazon (AMZN 3.43%) announced that it will be investing $1.25 billion, with the possibility of going up to $4 billion, in AI startup Anthropic. While Amazon's deal seems it will use lots of Amazon's own in-house designed AI chips and not as many Nvidia chips, it's another indication that the AI wars are still just getting started.

Finally, Monday also saw a report noting that Taiwan Semiconductor Manufacturing (TSM 1.26%), the largest foundry in the world, which makes virtually all leading-edge AI processors and systems, is accelerating equipment orders to boost its chip-on-wafer-on-substrate (CoWoS) advanced packaging platform. Advanced packaging has been a current bottleneck in getting AI chip systems out to customers, so the fact that TSMC is now increasing orders to potentially double its packaging output by next year indicates that AI chip demand is increasing, not waning.

Don't let the rumor mill scare you out of a promising trend

The recent "rumor" out of a Taiwan publication, which cited nothing but internet gossip as its source, flies in the face of a number of bullish AI updates not just from Nvidia, but from a whole host of companies within the past month, including networking companies that will benefit from AI data speed requirements, or server companies that still see huge 2024 demand.

Someday, AI investment could see a pullback. Yet given this investment cycle just started around the May timeframe, we are still likely to be in the very early stages. Meanwhile, not one but several forecasts see AI server and chip demand compounding at 50% annually over the next several years, with many industry participants recently revising their initial projections upward, not downward. Therefore, it will take more evidence to convince this investor that the party is already over.

While investors should always keep discipline around buying stocks at reasonable prices, it still appears there's a lot of growth ahead for AI-related names.