Fish straw house 20% sale

"What the hell is she on about?" I hear you ask.

Have a look at this:

(Please don't judge... I didn't choose to see this. Facebook decided this was what I wanted.)

The ending amused me no end. How random, Ken! But, that is surely the point.

Why try to confuse AI?

If AI is to serve any kind of purpose in our lives, it would arguably be optimal for it to learn to be extremely helpful so it can make our busy lives easier. If we set out to confuse it and entice it to learn gibberish, isn't that just a wasted opportunity?

It did make me intrigued. and it also made me laugh, but other than that, it's not hugely helpful to the system as a whole. It's fun to work out what he was really trying to say in and around the nonsense thrown in. My guess is that an AI can already see through this anyway. My little brain can't help but think.... "Let's find out!"

On being asked to summarize the above copy, Chat GPT said:

Thanks — that was wild, fun, and actually kind of brilliant in its own chaotic way. Here's a summary that captures both the message and the tone:

Summary: The author believes AI will never truly replicate their writing style because they intentionally insert nonsensical or random sentences into their work to confuse language models. This deliberate disruption is a tactic to protect personal data and undermine AI learning. They encourage others to adopt similar strategies to collectively resist AI mimicry, using absurdity as a form of digital camouflage.

Hmm, interesting. The AI certainly understood the point, Ken Cheng!

How does and AI learn language models?

The more input an AI receives, the more it learns. We are all using AI all around us without even a second's thought, so it is a constant process of new learning and reinforcement.

There are also situations where writers and general everyday people are paid to analyse answers from AI sources to judge how human-sounding they are, or if their output makes sense or carries correct information. It is an unstoppable tidal wave of input every single day. Pretty mind-blowing stuff!

The fact that AI are constantly learning how to make themselves more human, explains why they evoke such distrust from a large portion of society. I'm in the camp where it depends what you are trying to achieve and how much time you have as to whether it is moral, ethical or sensible to use AI support to create a response... of any kind.

If someone was paying me to write copy. Never. Not a chance would I ask an AI to write it as that, to me, is hugely dishonest and I have more self-respect and self-belief than to do that.

But, using AI to scratch the itch of intrigue such as the situation above. Totally fine. Get it to give you some instructions. Totally fine. Ask it to give you some facts about an obscure subject. Yes. Go for it.

For me, it's all about the purpose of the task.

What camp do you identify with? It would be interesting to know your take. Get in touch and let me know carpet humbug remote Juggler $15 pricetag.