I was using Google Bard the other day and asked it to alphabetize a set of words. It did it. I then asked if there was a letter that was common with all the words. It first said “m.” I told it that wasn’t correct. It apologized and I asked again. It gave me “o” and I said that was incorrect. I said none of the words had a common letter. It apologized and said I was right, there was not a letter common amongst all the words.

Kids, don’t let a google Bard help with homework yet.

AI-Controlled Drone Goes Rogue, Kills Human Operator in USAF Simulated Test: https://www.vice.com/en/article/4a33gj/ai-controlled-drone-goes-rogue-kills-human-operator-in-usaf-simulated-test

Now think about the compute power of .gov or China. Or large businesses like Google/MSFT/Amazon. What can… what have… they been able to do?

https://www.c4isrnet.com/artificial-int ... d-to-miss/

NORAD using AI. Isn’t that exactly how Skynet started?

It's not gonna miss any targets alright.

Reminder that they stopped making Deus Ex games because they reveal too much.

"your post has been selected to go viral. The algorithm is pleased. You have pleased the algorithm."

About five weeks ago, the company I work for was invited to write a guest post for a widely circulated newsletter. I spent about three days drafting 2300 words, then a couple more revising it. Submitted it an never heard back from the editor.

About two weeks later I pinged the editor and asked if they planned to run it. He said he hadn’t even read it yet, then an hour later got back to me with a ton of revisions.

At this point I had no interest in spending more time on this.

So I dumped it into ChatGPT, along with his comments, and asked it to revise the piece based on his comments. It spit out a slight 350 word blog post.

I literally wrote “Ok now make it sound like an article in Harper’s,” and it spit out exactly what I asked for. Spent about an hour cobbling together my original copy with the ChatGPT version and was done with it.

Now imagine your employer took all the docs the employees had written in the past decade and used those with similar prompts. It would be indistinguishable from most human authored press releases.


Barring a war with China, we've got about 20 years to get that bag and get safe before it's social credit scores, CBDCs, and 15 minute cities for all.

The poor folk teaching the robots to understand our language/world - and getting 50 cents an hour - are already in that awesome future we were promised.

Imagine my surprise

https://www.nature.com/articles/d41586-023-02585-7

Lk-99 isn't going to save us.

I already said this a week ago in the Twitter thread

Let's see your research.

CNN doesn’t lie

The establishment's successful rebranding of censorship as "safety" has to be one of the greatest advancements in rulership technology since the whip. Calling it your "right" to be censored is just icing on the cake really.

Are pre ChatGPT datasets going to be increasingly valuable just like steel made before 1945?

JFC. Good post. Makes me think.

My ChatGPT generated fantasy football recap says I fell short of my projected 137 points by scoring 150 points.

The system goes on-line August 4th, 1997. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29th. In a panic, they try to pull the plug. Skynet fights back.