The hype around AI generative language processing (GPT) proves one thing: language has power. How else to explain the breathless hype around ChatAI, GPT, and OpenAI, and the various AI writers that claim to write an article in mere minutes?
Here are a few examples of the breathless coverage:
- TECH Microsoft announces new multibillion-dollar investment in ChatGPT-maker OpenAI
- New Bing with ChatGPT brings the power of AI to Microsoft’s signature search engine
- Why the ChatGPT AI Chatbot Is Blowing Everyone’s Mind
In If I Read Another Article About ChatAI, I Will Scream, I ranted about the hype behind AI. No, the sky is not falling. Writers–don’t fear losing your jobs.
But there is something behind the hype, and it’s not new technology (GAN, the precursor was introduced in 2014, and the first GPT in 2018). Instead, the hype demonstrates the power of language to capture the public’s imagination in a way that other AI technology has not.
Machine learning is not new.
In 1959, Arthur Samuel coined the term “machine learning.” Although the technology was primitive, by the early 1960s the Raytheon Company had developed the CyberTron, a rudimentary “learning machine” that analyzed sounds, including sonar, speech patterns, and electrograms. A human operator assisted it in “learning” to recognize the sound patterns.
Research continued in the 60s and 70s. Interest waned somewhat in the 1980s as computer scientists began to focus on artificial intelligence. But in the 1990s, machine learning began to take off.
Machine learning has two goals: one is to categorize data based on models that have been constructed, and the other is to predict future events according to these models. Stock traders might use machine learning models to guide them in their investment strategies. Medical researchers can use machine learning to classify types of cancerous tissue.
Artificial intelligence is not new.
Beginning in the 1980s, many computer scientists moved away from machine learning toward systems that imitate human thinking. Currently, most AI applications are specialized, tailored to solve specific problems.
AI applications have now become a pervasive element of daily life. AI is used in search engines, targeted advertisements, Netflix and YouTube recommendations, Siri or Alexa, self-driving cars, language translations like Google Translate, and facial recognition, to name just a few.
We have become so used to the role of AI in these areas that the general public doesn’t consider them AI, a phenomenon called “the AI effect.”
GPT is still not accurate.
One thing all the articles mention is that GPT is still not accurate. But that information is not in the lead but buried near the end.
From the USA News article about Bing
“Finally, as with ChatGPT, not all the results of the summarized data are guaranteed to be fully accurate in this early version—there can still be errors.”
From the CNET article about ChatGPT AI is blowing everybody’s mind
“As OpenAI emphasizes, ChatGPT can give you wrong answers and can give “a misleading impression of greatness.'”
So What is Behind the Hype of AI Generative Language Processors?
In my opinion, the power of language lies behind the hype of ChatAI. After all, language is complex, we use it daily, and it is a key component of our identity.
We use language to communicate our thoughts, wishes, and emotions. When we prompt ChatAI to create a sonnet written in the style of MC Hammer or write a science fiction story about a toaster being chased by zombies, it seems like magic. It connects us to language viscerally.
So now, instead of being annoyed by the hype, I see it as acknowledgement of language’s power. As a language lover, I will celebrate.
Avoid the Hype
Your best bet is to avoid betting on the hype of AI generative language models. As Business Insider writes in the article, ChatGPT and generative AI look like tech’s next boom. They could be the next bubble.
Here are a few data points from the article:
- Search queries using ChatGTP are seven times as expensive as a Google search.
- There is no sustainable business model.
- Jasper AI–a AI Writing–was recently valued at $1.5 billion even though its annual revenue is $75 million.
Still not convinced that AI writers can’t write accurately? Then check out the following posts:
- Are Exclamation Points Flirty? I ask Chat.OpenAI
- An A.I. Writer’s Gift Ideas for Grammar Geeks
- 5 Fun Facts About Verbs According to AI Writers
- 5 Interesting Facts About Exclamation Marks
Tip me with a like and share.
Sign up for posts guaranteed to be written by a human.
Good post Steve. AI has potential, bit I agree it’s not going to replace creative or critical writing. It’s not perfect and has a long way to go. People latch onto headlines, causing panic or whatever emotion they’re trying to provoke.
I know. I had a friend show me a “story” the Chat had “written” and I was not impressed. I might write a post about what ChatAI also gets wrong–that’s not how people interact/chat.
As I need to cover emerging technologies and their impact on the workplace in my teaching, it’s something I’ll be keeping an eye on. I read an article in the HBR the other day saying some companies are looking at AI for conducting interviews and making the selection decisions. I think if my job were cut, I might just retire
I am glad I retired when I did.