What's different between ChatGPT & Siri? Will it replace us

Jan 20

ChatGPT, an AI chatbot, has been making headlines lately, replacing search engines, engineers, and workers. Why ChatGPT?

What is Chat-GPT?
To be precise, ChatGPT is a "big language model" built and trained by OpenAI, an AI research and development organisation founded by Elon Musk in 2015, which revealed the GPT-3 language model on June 11, 2020, and on November 30, 2022. Chat-GPT, based on GPT-3, launched on the 30th and allows users to register and chat for free on this webpage.

Language model? It is an AI that uses a lot of linguistic data to train to recognise and generate human natural language and accomplish its job using natural language commands. It speaks and understands human language.

ChatGPT has trained and processed 175 billion pieces of data on supercomputers to learn, understand, and translate human language.

When I started writing this article, I thought it would be better to let ChatGPT explain it to you directly so that it can be introduced quickly and easily and used to do some operations, like the above. I started to do some more complicated questions, and ChatGPT can answer in a very organised way in less than 5 minutes.

Why does "natural language" matters?
In the next query, I purposefully broke the text order and got a lengthy response from ChatGPT. Here, we can also extend ChatGPT's answer.

ChatGPT can even repeat topics. ChatGPT could interpret the dialogue and provide extra explanations or even correct its own language, just like a real human. I was writing a ChatGPT science post for you, but it seemed like interviewing someone.

First, we need a programme language to give a computer instructions, but the layman finds it confusing because machines comprehend language differently from people. Just like we don't say "if milk=0 else beer+1," life's discussions have more unclear space, inverted sentences, erroneous logic, and even some unstated tacit understanding, yet we can still readily understand them. ChatGPT too.

Will ChatGPT replaces all?
ChatGPT claims to help humans answer inquiries, manage tasks, and write content. Create news and comments? Isn't that what the editor's done?

In fact, similar topics are not new. Last year, someone used AI to paint to win an art creation award, which seems to break the imagination that AI cannot perform creative work. If you ask ChatGPT to write a poem, it can easily write a decent poem, showing that it is not much inferior to humans in text creation.

ChatGPT has been suggested to replace Google Search and certain programmers' work due to its use of natural language to solve problems and its enormous database content. Some of the descriptions in this page come from ChatGPT's own answers.

Will AI replace us?
AI can't feel emotions? In the above Q&A, AI combines "spring" and "sadness" to replicate human emotions. Though well-written, the poem shows despair. Will ChatGPT's poems get more accurate and emotional? The editor can't pack his bags and say it's impossible.

I watched a GQ magazine interview with Japanese robotics researcher Takayuki Furuta, and the exchange from 8:10 to 9:17 in the film answered these points nicely. Excellent invention breaks rules to expand our imagination. AI paintings and poems are based on past experiences, no matter how good they are. Picasso's spontaneous and bold creativity can help humans with data collection and aggregation.

So, back to how we don't want to be replaced by AI on this matter, I think ideas, inspiration, and experience are important factors that can make humans not easily replaced, because AI can't decide what to create today, AI can't write a sad love song because of a breakup, AI can't because of the care and understanding of people with disabilities to develop a variety of people with disabilities-friendly design, at best, we can only export these requirements.

Thus, ChatGPT emphasises practising our ideas, which may be the best method to start the AI era.

Finally: Active and Passive AI
In the last part, please allow me to take a bigger step to make the association, as AI is based on huge data and information to learn and create, their works are put into the broad sense that we feel "good" works to analyse the creation, which can easily let us lose more opportunities to contact new things. We now claim that social media platforms like Jitterbug and Youtube's filtering of recommended content to meet users' tastes produces a "filter bubble" that limits our knowledge to familiar and commonly accessed information. The network makes information access easier, yet it has limits.

In fact, this can also be seen in mainstream art creation, as Hollywood remakes ancient works to sell sentiment but is increasingly reluctant to invest in unique IP. If we continue to follow the AI creation model, we fear our creative, aesthetic imagination will be further confined. We're further from AI replacing humans when we're less exposed to new materials and stimuli.

So for AI, are we passively obtaining information from them, or can we take the initiative to use it for ourselves? The growth of AI may remind us that as our lives become more convenient and our work becomes easier, can we still use these simple chores as an indication of our self-worth? If you find it hard to roll with the AI, I can only say, resting flat or struggling to pick between the two is also the right to be a human being. I hope this post encourages the common.

Have you tried ChatGPT? Please a comment below!

All comments (1)
No contentNothing here, please try again later.