Technology

You know, it’s stuff like this that continues to irritate me to absolutely no end.

These dumbass chatbots you are referring to have already significantly changed the industry, for better or worse. The change has occurred. There are already so many things happening with these LLMs that are changing the way industries operate, and they aren’t even that good yet. Your personal bias against this technology has absolutely nothing to do with its applications.

The changes will increasingly be about computational power and the ability to process information at speeds and scales we have never seen before.

And this is where the potential of AGI becomes enormous.

Imagine an entity or device capable of taking essentially the entirety of human knowledge available to it, understanding the relationships between that information, reasoning across multiple fields simultaneously, and processing it at a speed and scale no human being could ever match. Now imagine that system being able to conduct research, identify patterns, generate and test hypotheses, design experiments, write software, analyze enormous datasets, and continuously improve its understanding of a problem.

That isn't magic. That is essentially the direction this technology is moving toward. Current LLMs are crude and limited compared with what a true AGI would theoretically be capable of, but they are already demonstrating the basic premise: machines can process and synthesize information in ways that are extremely useful to humans.

An AGI capable of reliably doing those things could provide an absolutely massive advantage to humanity. Scientific research that takes teams of humans years could potentially be accelerated dramatically. Complex engineering problems could be attacked from thousands of different approaches simultaneously. Medical research, materials science, energy production, climate modeling, mathematics, and countless other fields could benefit from having an intelligence capable of working continuously across enormous bodies of information.

That doesn't mean AGI would automatically make everything better. There are enormous risks, economic consequences, alignment problems, and questions about who controls such a system. Those are legitimate things to debate.

But dismissing the technology because you personally dislike current LLM chatbots doesn't answer the question of what the technology could become.

Go use the technology. Stop being a Luddite when it comes to LLMs. It is already being used as a tool. It already does things that are useful and makes many tasks easier for humans. I know that's an unpopular opinion on this platform, but it remains true regardless of how much you dislike the technology.

And, for the record, we have not fed "all of human knowledge" into an LLM. These systems are trained on enormous collections of available data, and there are legitimate and significant controversies surrounding the inclusion of copyrighted material in those datasets. Whether copyrighted works should be used for AI training without permission or compensation is a completely legitimate question. But that issue is separate from whether the underlying technology has useful applications.

The point is that AGI doesn't need to possess every piece of human knowledge to be revolutionary. It needs to be capable of acquiring, understanding, reasoning about, and applying information at a scale and speed that humans cannot match.

If we ever achieve that, the question won't be whether AGI can benefit humanity. The question will be how humanity chooses to use an intelligence with capabilities vastly beyond those of any individual human being.