What is ChatGPT?

Any good writer (or any writer in general for that matter) is taught to ensure that a written piece answers the 5 W’s: Who, What, Where, When, and Why?

Since AI is a vast and fast-moving topic (spoiler alert: I will be writing more blogs on AI), I thought it best to start at the beginning and address the question for the uninitiated: ‘What is ChatGPT?’

However, to add a bit of spice from the outset, I’ve written my version and asked the same question to ChatGPT.

So here goes, Man v Machine, Round 1…

1. Human version: What is ChatGPT?

“Our goal is to make AI systems more natural and safe to interact with.”

  • Source: Open AI
  • Dec 2015: Open AI was founded in San Francisco, initially as a non-profit organisation, with the founders pledging $1bn to the venture
  • Dec 2016: release of Open AI’s first software platform Universe to train its AI intelligence using the world-wide web
  • 2018: Elon Musk (yes, him!) left the board citing a potential conflict of interest with Tesla and its development of AI-technology for driverless cars
  • 2019: Open AI set up an for-profit subsidiary (OpenAI LP) and secured a $1bn investment from Microsoft
  • Dec 2022: free preview of ChatGPT (based on GPT 3.5)
  • Jan 2023: Microsoft announced a new $10bn tranche of funding, with an implied company valuation for Open AI of $29bn
  • 2024: Reuters sources predict Open AI revenue of $1bn for 2024

Glossary

  • AI: Artificial Intelligence
  • GPT: Generative Pre-trained Transformer (GPT)

So, in just over six years, we’ve come from a not-for-profit ChatGPT research prototype to a fully-fledged commercial AI application which values the company (Open AI) at around $30bn (and I’m pretty sure that figure is only going to go up…exponentially).

And what is behind all this wealth creation? A computer programme that allows you to type in any question and receive a considered answer written in a human-sounding way…and that’s the key part, human-sounding.

The fear is that ChatGPT (and other competing AI systems – there are quite a few around!) is so convincing, that it will do for the white collar workforce in this century what robotic automation did for the blue collar workforce towards the end of the last century.

2. ChatGPT version: What is ChatGPT?

ChatGPT is a large language model developed by OpenAI based on the GPT (Generative Pre-trained Transformer) architecture. It is designed to generate natural language responses to text-based prompts, making it useful for tasks such as language translation, question answering, and dialogue generation. ChatGPT has been trained on a vast corpus of text from the internet and other sources and has a deep understanding of natural language patterns and structure.

ChatGPT is one of the most advanced language models currently available, with the ability to generate coherent and fluent responses to a wide range of prompts. Its capabilities have made it useful for a wide range of applications, including customer service chatbots, personal assistants, and language translation services. With continued research and development, it is likely that future versions of ChatGPT and other language models will become even more sophisticated and useful for a wide range of tasks.

Analysis of the two versions

  • Human response
  • Attractive to read: subtitles, headings, pull quotes, italics, emboldening, tables…etc.

    These all serve to keep the reader’s eye ‘interested’ as well as highlighting key information.

  • Enhanced SEO: embedded into the text, and highlighted in H1, H2 headings.

    SEO is also enhanced by the use of quotes, tables, and captioned images.

  • More rounded argument

    A human response can look at the question behind the question and include additional information which will aid the reader’s understanding.

  • Computer (ChatGPT)
  • All text

    The text is split into paragraphs, but with no titles, sub-headings…etc.

  • No SEO: not purposeful.

    Note: SEO would be helping (inevitably) Google search rankings for a Microsoft-backed product. Just saying…

  • Critics cited: ChatGPT does take reflect opposing views – this is what can make it seem (at its best) as ‘human-like’.

Conclusion: Quality v Quantity?

So there you have it – which do you think is the better explanation?

Of course, the one huge advantage that ChatGPT does have up its sleeve is that its a computer which doesn’t get tired and works 365 24/7… so ChatGPT will always win on quantity, and if the quality is deemed passable/acceptable, then humans will be fighting a losing battle.

Related articles