Can AI be harmful to humankind?

Potential Risks

• AI can be programmed to do almost anything, including mass destruction. Artificial weapons can be created for the military. The potential AI arms race that could occur with the increasing level of artificial intelligence could result in the casualties of millions.
• Another potential risk of AI is incorrectly aligning the goals to our own. Meaning, the AI will attempt to accomplish it’s programmed goal at any cost, even if it is a destructive means.

Myths and Misconceptions

• One of the largest and common fears of AI is the worry that the AI will turn malicious or conscious. This is extremely unlikely and a more rational worry is that the AI’s goals are not aligned with our own, as mentioned earlier. This can lead to destructive tendencies and cause more harm over time.
• Another so-called “mythical worry” is that superintelligence is just a few years away. This is wildly inaccurate because the superintelligence that people are so fearing is at least decades away.
• Finally, another fear that is thanks in part to the many AI movies; robots. Robots are not a remotely rational fear because a misaligned intelligence does not need a body to do harm, all it needs is an internet connection.

Source: https://futureoflife.org/background/benefits-risks-of-artificial-intelligence/?cn-reloaded=1

Home

Table of Contents