China's version of ChatGPT... coming soon

This is the Synthetic mind, the Rachel Ray of Artificial Intelligence. This week weve whipped up a lunch that puts your grandmas biscuits to shame.

Here's how we're closing out the week:

  • China Set To Release ChatGPT Competitor In March

  • Hackers Using ChatGPT For New Cyberattacks

  • Quick Bytes

  • Mind Memes

China Set To Release ChatGPT Competitor in March

Apparently AI is the pizza that everyone wants a slice of. OpenAI, Google, Microsoft. Youll remember the good ol days when AI was just a race to see who had the better robot.

Now, those days are OVER.

The real games are about to begin: China will release its own version of ChatGPT in March.

Its like watching a food fight, but with money

Were told that Chinas version of ChatGPT will be more advanced than any other AI chat weve ever seen.

So whats their special ingredient you ask?

It's called, the Kunlun AI Chip. The most advanced chip for AI technology.

Let us explain:

  • Over the last few years China has invested a whole lot of money into AI

  • This money has been used to fund research projects like ERNIE (a Chinese language model) ERNIE has beaten BERT (Googles language model) and GPT-3 in most areas.

  • To slow China down the US stopped companies from sending AI tools to China

  • China creates the Kunlun AI chip

  • The Kunlun AI chip will make Chinas version of ChatGPT faster and smarter than Googles and ChatGPT

(Careful, youre drooling)

Wait?!... Sniff. Sniff. Do you smell something burning?!

China may have set the cook temp a little too high.

China is already finding it hard to walk the line between its censorship laws and teaching ERNIE. Language models like GPT-3, BERT, and ERNIE learn by talking with people.

So if there is a tight limit on what can be talked about then it will be hard for ERNIE to keep up with language models from other countries.

In short, China might be looking at one burnt pizza.

Hackers Using ChatGPT For New Cyberattacks

Over the last few months ChatGPT has been the perfect virtual sidekick. Smart, always available, and never takes your spotlight.

But what if our partner in crime was less like Robin to Batman and more like Loki to Thor?

What if we told you ChatGPT is teaching hackers the best way to hack into your bank account?

You heard that right. People are saying ChatGPT is teaching hackers how to hack you.

ChatGPT does its best to help everyone but it doesnt always know if what its telling you is harmful. Weve seen OpenAI publicly say this is one of their biggest struggles as a company. OpenAI tries to tag as many phrases or words as possible that suggest youre up to no good but圬rum roll please..

Hackers are finding ways around this by carefully wording their questions to trick ChatGPT into giving them answers.

Its rumored that ChatGPT has:

  • Taught newbies how to start hacking

  • Taught how to create a website on the Darkweb

  • Explained which part of a system/code is weak and gave the best way to break it (including banks)

  • Helped scammers by creating fake emails that look like they are from big companies (like banks). These emails are then used to steal private information.

The scary part about this is not that ChatGPT is teaching hacking but that it saves hackers a lot of time. Were talking hours to seconds.

This means that hackers will be able to reach more systems faster.

This is the part of the movie where Thor says Loki. How could you betray me?

But just as Loki saved Asgard, ChatGPTs advice may be just what these businesses need to save themselves. They are able to take this information and use it to prevent major data leaks, strengthen security systems, and create new systems that are almost impossible to hack.

Its funny how we all used to be scared of the big kid on the block. Now were all scared of the nerdy one with a laptop?! Oh how the tables have turned

Quick Bytes

Mind Memes

Join the conversation

or to participate.