AI is hungry for data. And it’s running out of food.

In order to make artificial intelligence more intelligent, it needs data. Lots and lots of data, quality data in order to train the system. Well, now it seems that quality data is running low. These hungry AI systems have gobbled up most of the quality data and they're looking for more.  

According to theweek.com,
“Trouble is that the data is running out. A paper by Epoch, an AI research organization, found that AI could exhaust all the current high-quality language data available on the internet as soon as 2026. This could pose a problem as AI continues to grow.”  

Massive amounts of data. In order to train ChatGPT they wound up using approximately 570 gigabytes of data. That's about 3 billion words. A lot of the stuff that we find on the Internet is totally useless for creating AI models. The Wall Street Journal says.
"Companies also are experimenting with using AI-generated, or synthetic, data as training material — an approach many researchers say could actually cause crippling malfunctions." 

google is considering going internal with google docs and Google Sheets. Other companies are looking for content that can't be found on the free Internet by going to large publishing companies. Meta has considered buying a company like Simon and Schuster for their content. (Authers might have something to say about that) 

The other option is to reconstruct their learning algorithms. They might start to let AIs gather lower quality data or possibly move to smaller models that require less data. 

It's a brave new world out there, and right now there's no way to tell what direction it's going.  

I'm all for artificial intelligence. As long as it doesn't take my job. 

AI is running out of data | The Week
The case for and against giving AI the kinds of rights humans have | The Week