• Writing Inc
  • Posts
  • šŸ¤– How the US Military is already using AI

šŸ¤– How the US Military is already using AI

PLUS: How Bing AI already lied to us

Happy Valentine’s Day.

Here’s something to end it on.

Source: Reddit

On to our stories.

Air Force lets AI take the wheel…sort of.

This week the US Air Force revealed an AI piloted one of its F16 fighter jets for about 17 hours.

One step closer to Terminator becoming a reality.

It took over for several takeoffs and landings, the Air Force said.

While this isn’t surprising, it’s a step into uncharted territory.

Source: US Air Force

As the military digitizes, decision-making could be more automated. Autonomous drones proved themselves as effective weapons for the U.S. military. While they’re still operated by a human, the tech is there to make them completely autonomous.

Companies like Boston Dynamics are developing combat-ready assistants to soldiers.

AI cannot drive war though.

A Soviet soldier saved the world from annihilation thanks to this. Vasily Arkhipov famously refused an order to launch a nuke during the Cuban missile crisis. AI can't make those decisions.

At least, not yet.

ChatGPT human after all: made up some answers during demo

Awkward.

It took a while, but someone finally noticed that a few answers from Bing AI’s demo last week were, well, made up.

Don’t worry, Bing AI. We’ve all been there.

Bing AI completely made up some numbers when asked to compare earnings results for Gap and Lululemon. Other calculations were off at best.

Source: DKB Blog

This sort of thing isn’t new. ChatGPT also makes stuff up. A lot. It will cite papers that do not exist, and events that didn’t happen.

ChatGPT isn’t so much a factchecker as it is a generative text AI. So, sometimes, it’s not generating the correct answer, but something that sounds right.

This is a point the major companies investing in these technologies emphasize, repeatedly.

It’s also a tough egg to crack. ā€œWhen was the Battle of Waterloo?ā€ is a question with a factual answer. ā€˜Why did the U.S. pass the Patriot Act?ā€ has more nuance.

We expect these AI chatbots to answer both, right?

Video game developer faces backlash over AI generated image

Uh oh. Looks like not all art is created equal.

At least, not AI-generated art.

The video game developer Bungie took some backlash this week over AI-generated imagery.

They showcased some fanart as part of a weekly contest, but didn’t check things closely enough. Fans discovered the art was actually AI-generated.

See for yourself.

The user who uploaded it shrugged and said 

ā€œWoah, I just thought the picture was really neat so I posted on the creations page. I’ll take the post downā€

Typical.

This was Bungie’s response.

Last week, an A.I. art submission was mistakenly featured in our blog. The process of choosing these involves a team effort and with this technology being so new, we don’t have a foolproof way of knowing what submissions are A.I. art.

We want to keep this celebration of our community for those that work hard to bring their creative selves to the forefront when creating works that the Traveler would find joy in. Because of this, we will not knowingly ever feature A.I. art submissions as a potential #Destiny2AOTW or #Destiny2MOTW winner. That being said, this is still new. We ask for grace if we mistakenly feature a submission generated by A.I., and a respectful heads up should it ever happen again in the future. Appreciate the assist!

Bungie Blog

I do not envy their position. Unless you’re looking for the six-fingered man, it can be tough to spot real from ā€˜fakeā€.

Westley realizes that he’s talking to an AI-generated person.

There doesn’t yet exist a great tool for detecting this.

What this further highlights is the backlash creators send towards AI art. For casual observers, there’s no controversy.

But many creators worry about the effect AI-generated art will have on their livelihood.

For now, the community must act as the sleuth.

🤯 Linkin Park returns…as an AI-generated music video?