Exclusive: IBM researchers easily trick ChatGPT into hacking

[ttown]

That’s the dangers of AI, new laws will have to be written so pure evil isn’t unveiled upon the nation IMO


https://www.axios.com/2023/08/08/ibm-researchers-trick-chatgpt-hacking

 

[conditionzero]

Oh you sweet summer child... What is done is done. The only laws that would be written, if ever, would be to protect the federal governments use of AI. Not to protect the citizens.

 

[Chief Sapulpa]

That’s the dangers of AI, new laws will have to be written so pure evil isn’t unveiled upon the nation IMO


https://www.axios.com/2023/08/08/ibm-researchers-trick-chatgpt-hacking

Evil doesn’t abide by laws or moral ethics.

 

[ttown]

Do we not have product liability on guns and cars? Software put out by companies that haven’t provide adequate testing and safeguards could very well be included IMO.

If companies are held libel instead doing product testing on the public you’d see a whole different attitude.

As far as laws many computing laws we’re put into place in the 70’s that were very effective when none existEd.

We had a phd scientist writing code for a major collaboration between computing research and IBM placed code in the program that slowly added bugs and erased itself if his name was removed from the HR database. Backups became corrupt as they were trying to resolve these issue. It took at least two years to resolve the problem.

This is no different IMO, glad I didn’t have to work with a bunch of negative Nancy’s that simply gave up. We were results oriented.

As technology advances new laws will be required…..