I haven’t written much about AI since my post in December went viral, but I collected enough to post something else.

I want to start with a meme that someone posted on Facebook and it pretty much says it all.

Sam Altman Desperately needs you to believe that generative AI will be essential, inevitable and intractable, because if you don’t, you’ll suddenly realize that trillions of dollars of market capitalization and revenue are being blown on something remarkably mediocre. If you focus on the present — what OpenAI’s technology can do today, and will likely do for some time — you see in terrifying clarity that generative AI isn’t a society-altering technology, but another form of efficiency-driving cloud computing software that benefits a relatively small niche of people.

If you stop saying things like “AI could do” or “AI will do,” you have to start asking what AI can do and the answer is… not that much, and not much more in the future. Sora is not going to generate movies. It’s going to continue making horrifying human-adjacent creatures that walk like AT-AT’s from Star Wars, and cartoons that look remarkably like copyrighted material from YouTube.

This is pretty much sums up my thoughts, too.

Imagine spending hundreds of billions of dollars creating an app that writes books and generates art. But all the books it writes are lame, and all the art it creates is derivative, so the only actual use people find for this technology is to assist writers with new writing tools, and assist artists with new art tools that probably would have come about eventually without the hundreds of billions of dollar investments in this tech.

We were blinded by sci-fi. Some like cell phones caught on. Some like VR goggles did not. Some like Second Life and Google Earth were popular for a while, but then we exhausted their capabilities and moved on. Some like driverless cars turned out to be a lot harder than we thought it would be. The entire AI industry is stuck in Sci-Fi delusion mud.

AGI is a lie, it wont happen. Artificial Impersonation is my term for what the world is calling “artificial Intelligence”, because I don’t consider computer algorithms to actually be intelligent. AI systems write by impersonating millions of writers scooped online. AI systems make art by impersonating millions of artists works scooped up online.

What the AI industry wants is AGI, or Artificial Einsteins that are smarter than most humans, and there are not millions of Einsteins online to scoop up. Here is further evidence:

GPT-4o is not a huge improvement over GPT-4

Let’s start with the biggest news. OpenAI released a new Chat GPT called GPT-4o, that’s a lower case O not a zero, it stands for “omni” whatever that means. Of course all the AI Bros were all over it, singing the praises of OpenAI, comparing it to Samantha from the movie “Her” and other over representation.

Because the OpenAI marketing video looks like a staged marketing video, I was waiting for someone to actually test it out with some skepticism. Toms Hardware had this to say:

Using GPT-4o it is instantly clear how much faster it is than the earlier models, including GPT-3.5, which is much smaller and less capable. It can also analyze video content, which is something not previously possible in ChatGPT or any mainstream chatbot.

Tom’s Hardware “I gave 5 prompts to ChatGPT-4o vs GPT-4 to test the new AI model — here’s what happened”

So, it is faster. Not a surprise, OpenAI is using the latest and greatest AI chips from NVIDIA which is currently the most valuable company in the world just because they make AI chips. But is it smarter? GPT-4 came out over a year ago, they have been improving the model with more and more questionably sourced data. Is it actually smarter? Well no not really.

Conclusion

I don’t think GPT-4o Omni is a significant step up in reasoning capabilities over GPT-4 but it is more descriptive, faster at responding and its big differentiator isn’t text but multimodality.

ibid

A year of research has made GPT faster and maybe more user friendly, but not noticeably smarter. That’s because it can’t be.

The Truth About AI Video Playlist

I wanted to write more, but it is all said in this You Tube Playlist I set up. First video is a good summary of current state of AI. Second is about how bad the financials are in the AI industry, Third is a scientific exam of the mathematics behind Large Language Models, and why they will never get smarter much beyond where they are currently at.

After that are some random AI videos about the problems and general lameness of the industry. I’ll keep this playlist up to date when I find more relevant videos.