ChatGPT and Generative AI

We're on the verge of the next industrial revolution; the next paradigm that will change how the developed world operates.

Generative AI; best known in the recent release of ChatGPT.

What is ChatGPT?

ChatGPT is a Large Language Model (LLM) chatbot that can answer your questions, and even right essays for you (hence why it is 'generative'). The 'GPT' part stands for Generative Pre-Trained Transformer.

Why is there so much hype around ChatGPT?

Ignoring the various online assertions that ChatGPT can do your homework for you (which may be true at a basic level) , it is indeed a ground-breaking development.

The chatbot could replace Search as we know it (eg: Google search) by becoming a more advanced personal assistant (think Siri on rocket fuel); a personal assistant than can write your Python script for you for your next trading bot.

Whilst Google Search will tell you where to look and find your the answer to your request, ChatGPT will tell you the answer direct. I have no doubt that applications such as this will overtake traditional search in the not-too-distant future.

But how good is ChatGPT reeeeaally..?

We must always remember that despite how slick the user experience is with ChatGPT, and the eloquence of the response, the information that we receive can only be based on what the AI has been trained upon.

It's not magic.

It's not a seer of the future.

Its answers cannot yet be fully trusted without some sort of human verification.

But nevertheless it is a massive step forward.

Microsoft vs Google

Having worked in the AI space for sometime I always saw Google as the Mac Daddy OG of AI, particularly after their acquisition of Deep Mind back in 2014.

I (naively) never thought I'd see the day that anyone could get in front of Google when it came to the AI field.

I always considered Microsoft to be a wannabe in AI. The spotty kid standing in the corner at the AI dance, looking on wistfully at the belle of the ball, Google, dancing around the room courted by its hoard of admirers.

I saw it in real terms when we were trying to hire data scientists on what could be considered a good engineering salary for a start-up - and finding the conversations ending quickly when we were told that Google had offered them the same job at $200,000 per year. It happened time and time again.

We could not compete with Google.

It felt like nobody could beat Google in the AI space.

But some time ago Microsoft made a masterful move; it invested in OpenAI, the company behind ChatGPT - which gives it a privileged position in the use of its technology, including ChatGPT.

So that spotty kid at the AI school dance is now becoming a handsome and beautiful* leading authority within the AI space, whilst Google is finding itself on the backfoot needing to do something to show that it's still the leader of AI - not helped by the embarrassing mistake made by Bard, its ChatGPT equivalent recently.

Are other companies involved in Generative AI?

Oh yes. They have tough competition, particularly with Microsoft who are now doubling down on AI in their commercial sales, but there are plenty of smaller companies and start-ups rapidly working in the space.

One such company is Aleph Alpha; a German start-up hoping to find a niche for itself in the big Generative AI space.

One unique factor to consider is about data privacy.

If you're working with Microsoft on a Gen AI solution (or anyone else in 'Big Tech') then it would almost certainly be hosted on their own Cloud platform; Azure, in Microsoft's case.

And the same with Google on its GCP platform.

These don't allow much privacy within those Big Tech ecosystems, and other companies may be able to provide more private options, perhaps bespoke on-premise solutions to keep the data truly private.

Good luck, John Connor

Whatever will happen in this innovative new space, its certainly early days and it will be exciting to see how this technology unfolds over the coming years (and even the next hundred years), and which companies will become its main protagonists.

Notes and caveats

*Yes agreed - appearance and looks are not important. I only use the phrase for metaphorical merit.