Once again, we have to assure you.

This article is 100% human-written.

So many fake people these days.

🤖 Top Stories

😬 Google might have swiped data from ChatGPT

Two big rumors from Google AI.

One: Google AI and DeepMind are uniting against a common enemy: OpenAI.

Even though they’re both part of Google, they're fierce rivals. That’s peak “enemy of my enemy”.

Two: Google trained Bard, its ChatGPT competitor, using data FROM ChatGPT.

That’s…a yikes.

Allegedly, a former Google researcher spoke out against Google using data from ShareGPT.

That’s a website that scrapes data from ChatGPT.

The rumor is the researcher quit to go to OpenAI. Google then scrubbed that training data from Bard.

Whatever the case, this type of rivalry between Microsoft and Google is refreshing. The old companies still have some fight left in them.

👀 Microsoft debuts another copilot, this time for security

Microsoft paid ALL THIS MONEY for GPT-4, and it’s going to use it, darn it.

First it was Github copilot. Then it was Office 365 Copilot. Now, it's a security copilot.

Well, I guess first it was a literal copilot, because they have a flight simulator.

Security copilot is a security assistant designed to, well, copilot with security professionals. It helps them identify incidents and trove through the mountains of data they deal with.

It’s a natural language assistant, so expect to be able to ask it about your security incidences, or for summaries of events and vulnerabilities.

Incidences and attack angles can turn into powerpoint slides thanks to the assistant. Prompts bundle into little functions that you can share.

In short, a security professional’s best friend.

Now, let’s hope no one hacks it.

🤓 UNESCO says be careful with AI, please

Another one. The United Nations joins the growing list of organizations and individuals concerned about AI.

The United Nations Educational, Scientific and Cultural Organization, or UNESCO, published a paper that says the following.

Listen, guys, we have to pump the breaks on AI development until we can guarantee two things. One, make sure AI is working FOR humanity. Two, make sure that it follows the law. And three, make sure your country checks in with us and follows our guidelines.

Beyond sanctions, there’s not much the UN can do to prevent, say, Iran from developing AI outside of their rules.

It does, however, show how powerful AI is. In the wrong hands, it could be even worse.

😎 AI tools, tips, tricks

1. AI game art → pixel art. Learn more

2. More prompt engineering resources. Learn more

3. Host ChatGPT locally. Learn more

4. More generative AI for consumer goods. Learn more

5. Turn your prompts into templates. Learn more

AI got jokes.

Source

🤖 Got feedback for us?

Head over to our Twitter, @RobotRockDaily, and let us know what you’d like to see more or less of.