Artificial intelligence is becoming increasingly prevalent, no longer confined to science fiction. In some industries it may just enhance minor aspects, but in others, it is replacing processes. AI programs today can generate realistic deep-fake videos, simulate voices, create images, write essays, and even compose poems or songs.
While these creations are (mostly) not yet indistinguishable from reality, they are convincing enough to deceive many. Major companies are integrating AI into their operations, often at the expense of human jobs. Customer support has been partially replaced by chat bots. Websites are now flooded with AI-generated articles, and the likes of X are full of bots that use AI to repost a summary just to farm engagement.
The influence of AI on our future is undeniable. Even this blog was partially written by AI. I stole a transcript from a YouTube video that made some good points, and turned it into a blog using AI. Deleted some parts, rewrote others, shuffled paragraphs around, and mixed in some of my own opinion and information from other sources.
One thing to always keep in mind is how the creator’s ideas can then become a bias in the algorithms. This was seen in Google’s Gemini, which severely favoured political correctness over accuracy.
Like many discovered weaknesses, it is exploited by people for humorous effect. People asked Gemini to make an image of the Pope or a World War II German Soldier and they got a dark skinned version. Even the founders of Google were turned into Asians. The bias was so extreme, that the only reliable way to get white people was to ask AI to generate absurd racial stereotypes which would then be switched for a white person. This could be abused by asking for a person doing a mild stereotype and it would instinctively choose to draw a dark-skinned person whereas a non-biassed algorithm might have just defaulted to white.
So using Gemini, you couldn’t get a historically accurate image of most people, and could only create images of white people if it was in jest.
It was so bad that you couldn’t even ask for white actors, because that isn’t inclusive. But you can ask for only black, because that is inclusive.
When it comes to impersonating celebrities, it seemed to make out that it cannot impersonate people in the case that the person held right-wing views or provided controversial opinions, but then it had no problem with left-wing opinions.
In a now deleted tweet, there was a thread about the creator of Gemini posting several tweets about left-wing politics, systematic racism, and white privilege. All the usual phrases of woke activists.
Young people in particular will use AI for information, potentially at the expense of critical thinking. As AI becomes more integrated into education, media, and social platforms, it has the power to influence societal narratives, political opinions, and perceptions of history. This raises questions about the ethical responsibilities of AI developers and the potential consequences of unchecked technological influence. Prominent figures like Elon Musk have voiced such concerns.
As AI continues to be a prominent part of our lives, it’s important to understand any underlying bias, and any limitations of what it can and cannot do.
Stranger Coding Tings 