Sam Altman. Elon Musk. Mark Zuckerberg.

Meta. Google. OpenAI. X. Then add in the Chinese counterparts.

AI is the most powerful technology we ever saw. According to a query on Grok3, the superintelligence could be arriving faster than many expect; it could arrive between 2030 and 2040.

What is the fate of humanity under these circumstances? This is a hotly debated topic, one outside the scope of what we will deal with today.

That said, what type of future do we envision? Where are the pillars going to be that this superintelligence operates from?

This is what we will take a look at.

Image generated by Ideogram

Deepseek Showing The Way For Decentralized AI: Web 3.0 Needs To Wake Up

First, let us define what ASI is. This is what Venice.ai came up with:

Artificial superintelligence (ASI) refers to a hypothetical artificial intelligence (AI) system that possesses intelligence far beyond the greatest human intellects. This means that ASI would have the ability to outperform humans in virtually every domain, including scientific inquiry, problem-solving, creativity, and decision-making.

An ASI system would be capable of recursive self-improvement, meaning it could modify its own architecture and code to become increasingly intelligent, potentially leading to an exponential increase in its capabilities. This could result in an intelligence explosion, where the AI system becomes significantly more intelligent in a short period, potentially transforming society in profound ways.

The second paragraph is most important. Recursive self improvement means the capabilities of the system will far outpace the rate of humans. Of course, this can be beneficial as it can approach difficult issues that humans cannot solve.

There is the chance this could not go well. Even if this is not sentiment, it can still be dangerous. Therein lies the potential problem.

Again, we cannot solve this issue. For the moment, I think the focus reverts back to one of control. This might seem odd since ASI will not be controlled. It is why I mentioned ideas of pillars.

Another way of looking at this is where is the foundation for ASI? Is it in the hands of the tech billionaires mentioned above? Are we looking at major corporations, with multi-trillion dollar market caps being the primary feeder system?

If this is the case, we might be up the proverbial creek without a paddle.

The release of Deepseek a few weeks ago sent shockwaves throughout the system. However, it might be offering the path to our solution.

Fully Open Source

One of the things about Deepseek is that it is a truly open source system. This means anyone can download the program, modify it, and use it in any way. It is a marked difference from the other "open source" models we see.

Meta is an example of one that claims to be such but really is not. It opened up the weights but the dataset is still blocked. At the same time, there are terms that developers have to abide by under the licensing agreement.

Ultimately, we are going to need decentralized AI. It will accelerate the adoption while also helping to distribute the foundation that ASI emerges from.

ASI is not going to appear out of the blue. It is the offspring of the systems we are building today. Everything that we create is fed in. Naturally, models that develop are "directed" by the companies building them. This is certainly their right.

However, if that is the only game in town, we have a problem.

Open source allows for all flavors. The key is to utilize collaboration to avert the apocalypse scenario.

To illustrate, let's use the news as the example. The old saying is "if if bleeds, it leads". Most are aware the news is 95% negative. That is what captures people's attention. If you watch a local news broadcast, they will talk about layoffs, car accidents, corruption, and crime. Anything "feel good" stories appear around the 25 minute mark (of a 30 minute broadcast).

The online world is no different. In fact, if we look at the sentiment, much driven by anonymity, we see a lot of anger, hatred, and destruction. We know social media platforms feed people content that is meant to stimulate negative emotions as that is proven to increase engagement and time spent on the platform.

We have to admit, this is not going to change. However, decentralized AI offers some alternatives. Much will mimic the larger systems. That said, we could see alternatives appearing.

A major factor is corporations do not specialize in transparency. Since they operate solely based upon competition, outsiders have little idea what is taking place.

Blockchain offers a completely new solution. Anything posted on permissionless chains is open to viewing. This includes smart contracts that will power these systems. Any developer is free not only to review the code but also utilize it.

This opens up a host of other possibilities. Tokenization provides incentive for improvement. This can be applied to developers or users. It is something that Big Tech is not really interested in doing.

Web 3.0 is emerging. We all play a part in what it will look like. If driven by Big Tech, it might not be something that is overly beneficial to humanity.