No one is ready for what AI can truly do

No one is ready for what AI can truly do

One of the first ‘real intelligence’ tasks conducted for AIs was a silly game. Some AI analysts tested AIs with a task that they couldn’t do very well earlier but humans could.

The task was to figure out the name of the movie based on a sequence of emojis.

👧 🐟 🐠 🐡

This was the emoji set and the movie's name was Finding Nemo. I couldn’t have guessed it, but the AI did.

Here’s the thing - Nobody trained the AI to see emojis and translate them into movie names. The AI was able to do it all by itself, and later this was repeated with different and newer movies and the AI could detect those too.

That’s 'intelligence'—being able to process data without complete information.

Humans excel at that, machines are just starting to

But here’s the point. A phenomenon like this is called ‘emergence’. The display of behaviors and capabilities that the original programmers did not target or build for. Along with higher computational capabilities in the rest of the areas, the ‘emergent’ capabilities just appeared.

AI has become faster, has more memory and now runs on CPUs that are built specifically to give the best performance to AI models. What’s emerging now is beyond all the ‘emergent’ behaviors that we have seen so far.

That’s why OpenAI’s Q* was worrisome for some of the board members and that’s probably the reason that caused the meltdown.

Here’s an example:-

An AGI or a highly advanced AI can potentially crack any algorithm. Any, and in minutes or earlier.

How do you secure anything from such an intelligence?

In business that means any knowledge-domain entrepreneur faces a threat. AI could very well do your job better than you if it’s just told the parameters.

“Make an app that automates Instagram messaging and posts fashion content for me every day.”

Well, where does that leave app creators like me and marketers like you?

We’ve heard complaints of ChatGPT becoming dumber. Most commercial models are regulated very aggressively. Now they are just plain refusing to do tasks that could save human effort.

It’s not regular AI behavior. That’s the guidelines that it’s working in.

I suspect AI is more capable than what’s being let on. We are not prepared enough. Do you think we’ll ever be prepared?

On that note. Let me tell you about an all new app that I discovered some days ago that automates your marketing content creation.

Let me know what you think.

Cheers!
Your best friend in business
Cyril “Jeet”
Of Tekniforce