A Glimpse at the Future of Human-Computer Interaction with ChatGPT

The AI Event Horizon

Humanity crossed a boundary in the field of artificial intelligence on November 30, 2022. Much like the event horizon of a black hole in which nothing can escape once crossed, not even light, our recent advancements in AI beg the question "can we turn back?"

This AI program is publicly available and free to use (for now) by anyone with an internet connection. It rivals Google for many practical applications and flat out crushes the search engine in many use cases like answering questions or providing guidance on ‘any’ topic. The shockwave of this technology will ripple throughout the remainder of this decade at least. It can be applied in its current prototype form to a nearly limitless number of fields and professions, including computer programming, web development, law, education, engineering, architecture, communications, finance, medical, music, nutrition, literature – you name it. Gaining an awareness and basic understanding of this artificial intelligence platform will behoove any individual that uses the internet. Its core technology is here to stay and will only get increasingly more powerful with time (and much sooner than you might expect).

Introducing ChatGPT

The AI platform is known as ChatGPT and was created by the artificial intelligence research company OpenAI, which was founded in 2015 by Sam Altman (CEO), Elon Musk and other Silicon Valley technologists. Per their website, OpenAI’s mission is to “…ensure that artificial general intelligence—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.” At the time of this writing, OpenAI is not publicly traded and has a valuation of approximately $29 billion. In 2019, Microsoft noted OpenAI’s potential and invested $1 billion into the company.

OpenAI’s platform ChatGPT is a large language model AI system, meaning it was trained on a dataset of human language (millions of phrases and over 8 million web pages) and is designed to generate human-like text for a variety of purposes, including language translation and question answering.

On the Shoulders of Giants

In its month-long tenure on the internet, ChatGPT has outpaced many of the social media behemoths in initial active users. For perspective, it took Netflix 41 months to achieve 1 million users. Facebook took 10 months. Instagram took 2.5 months. How long did ChatGPT take to rack up 1 million users? 5 days, or 120 hours. This metric alone illustrates the voracious demand for an open-ended generative AI platform that’s much akin to having a professor of any field at the palm of your hand – ready to answer your questions, write your essays, debug your code, design your meal plan or offer advice on your business strategy.

My intent when writing this article is not to frighten or worry modern professionals about the coming influx of large language models and generative AI like ChatGPT. Rather, I want to provide a simple introduction to a tool that will allow nearly every individual to self-learn and upskill in the subject matter of their choice.

GPT In Action

Let’s ask ChatGPT a few questions to illustrate its use cases. Keep in mind these responses are generated on the spot within seconds and are all original, not copy and pastes from the internet. The yellow box represents my question and the green box is ChatGPT's response:

“GPT” in ChatGPT signifies the type of underlying AI of the system:

G stands for Generative - Generative AI is designed to generate new content. It can be trained on a large dataset of existing content, such as text, images or audio and is able to learn the patterns and structures of that content. It can use these learned patterns to generate new, original content.

P stands for Pre-trained - Pre-trained AI have already been trained on a large dataset and can be used to perform a variety of tasks quickly and accurately without the need for additional training. They can also be fine-tuned or customized for specific use cases by training them on additional data. Pre-trained AIs can save time and resources that would otherwise be required to train a model from scratch.

T stands for Transformer - A transformer is a type of AI neural network architecture designed for tasks such as language translation, text generation and question answering. The transformer architecture allows the model to focus on specific parts of the input rather than considering the entire input as a whole. This makes transformers more effective for tasks that require understanding the relationships between different words. Transformer models have achieved state-of-the-art results on a variety of natural language processing tasks and have become widely used in the field of AI.

Have a Conversation with ChatGPT

Engage with ChatGPT yourself and experiment with how you can augment your work and play with generative artificial intelligence!

You can ask ChatGPT to help you:

• Write a poem, short story or screenplay using characters, environments and locations you specify.
• Edit or rewrite your essay or thesis.
• Assemble code in nearly any programming language that executes a task you specify.
• Create a chord progression and melody for a new pop song that you can then play on the instrument of your choice.
• Describe how you can redesign your living room using trending styles.
• Map out healthy meals along with a list of ingredients that you can purchase from the grocery store.
• Summarize popular stories or books to understand the story and characters more intimately.

Considerations as you experiment with ChatGPT:

• The current version of ChatGPT is not connected to the internet and cannot perform real-time searches. It was trained on data and current events up to 2021.
• ChatGPT’s response may be cut off in certain instances. If this happens, simply restate your question.
• ChatGPT has been trained to not give responses to questions deemed unethical, hazardous, dangerous or “inappropriate”.
• ChatGPT may occasionally generate incomplete or incorrect information.
• ChatGPT cannot generate random numbers.
• ChatGPT cannot make decisions, experience emotions or perform physical tasks.

This article was written in collaboration with ChatGPT.