There are some who are opining that they recent crash in crypto is tied to the view that the AI hype cycle is going to burst. For some reason, many are equating what is occurring now in AI with the Dotcom era.
I have discussed this idea on a number of occasions. The problem is we are dealing with apples and organges.
More on this in a bit.
Did the AI hype cycles, specifically fears around it, crash the crypto market? My view is that crashes are when sellers far outnumber the buyers. I have watched markets for decades and it seems the talking heads in the media (and online) all seem to think they know why markets moved in a particular direction.
The problem is this never matches up.
I recall one instances where one of the financial news programs was claiming that equities crashed that day due to oil going down. When I looked at a longer term chart, it was evident there was no correlation to stocks and oil for the last 6 months. In fact, they were heading in opposite directions.
So why did they suddenly correlate that one day? It is all nonsense.
Market Crash Tied To AI Hype Cycle?
To start, let me state there certainly are risk-on/risk-off events in markets. Crypto is definitely, contrary to what Bitcoin maxis assert, a risk-on asset. When investors have more appetite for risk, they will flood into crypto. Naturally, we see the opposite happen when fear starts to take hold.
Because of this, there could be some correlation to AI hype and crypto. However, we cannot reach any conclusions in a single day. Plus, we have to look at the totality of capital flows. Did US Treasuries, a safe haven, see a spike? How did the USD fare, another asset people acquire during fear stages?
Another factor is what happened with the different indexes. What did the Dow do compared to NASDAQ or S&P500? The former is a reflection of international capital flows whereas the other two are more US (and tech) based.
We can take a look at this situation, if correct, as a way to read into where things might be heading.
If crypto is reacting to the AI hype cycle, are we going to see a Dotcom era crash?
AI Is Not Hype
People are extremely shortsighted. They also tend to move in unison which is the majority never make any money.
AI is not in a hype cycle. The idea that we are entering an era of excess when it comes to this technology is absurd.
As stated in the past, the Internet era did not have a great deal of utility in the early days. Companies like Cisco were expanding bandwidth without saturation. This is not the case with AI.
Here is a quote from an article that talks about the bubble. Actually, it was the headline"
Experts Tout Decentralized AI Efficiency Gains as GPU Shortages and Energy Limits Loom
If we are going to see a shortage of GPUs, how are we dealing with a bubble. Companies like Nvidia will always expand their production. Where have we ever seen a company viewed as negative when it has overwhelming demand that cannot be met immediately. We cannot presume the company will expand at the same rate as the past but more GPUs will be output next year as compared to 2025.
Then we have the technology curve. Each new generation of GPUs has faster processors, is more energy efficient, and have greater storage.
It is Moore's Law but on steroids.
The key here is inference. To believe that more prompts will not occur in the future is absurd. Simply look at how much is rolling out. It matters none what is rolled out in terms of model training. Inference is where the demand resides.
Heightened circumspection, particularly following China’s Deepseek success, which shifted market attention eastward, has focused a critical light on Silicon Valley’s financials. The core concern now revolves around the evident disparity between ambitious, long-term revenue projections and the highly inflated, speculative valuations commanded by AI firms. Critics say these metrics suggest a significant correction may be overdue.
This highlights the crux of the issue. People are looking at revenue speculation when the reality is the major players in the industry are killing it.
Sure, we can focus upon OpenAI and cite it as an example. The issue is the major players include Google, Meta, and other Big Tech companies. Have you looked at the revenue streams they have in place?
Monetization is a problem similar to any new technology. That said, these companies learn how to leverage their innovations for profits. It is likely to only increase as real world AI starts to take hold.
Do you see a use case for humanoid robots? This is something that will emerge in the next few years.
If crypto is tied to AI, then we could be in for a longer bull run then the doomers realize.
Posted Using INLEO
This post has been shared on Reddit by @uwelang through the HivePosh initiative.
Yes as personal assistants 😊
Solid point on GPU shortages vs bubble dynamics. Real infrastructure constraints + growing inference demand suggests we're nowhere near dotcom territory. Revenue streams matter more than hype narratives.