Assistant or Replacement
Chatbots are a hot topic of conversation these days. Many colleges are concerned that chatbots are writing the papers that students are submitting for their grades. Some employers discovered that certain remote workers hold multiple full-time jobs, using chatbots to generate the research they deliver, allowing them to collect salaries from different employers for their time.
Chatbots are computer programs that use artificial intelligence (AI) to learn patterns of our information world, as used books, websites, journals, and other text materials. They have learned how we use words and phrases to express ideas. Chatbots accept text input and generate text output.
I use them as research assistants. Here is an example of an interaction I had with ChatGPT. It readily responds to my vague question.
I want to know how many people died in whatever than year was when some influenza thing struck the United States?
ChatGPT handled the uncertainty with aplomb:
There have been several influenza pandemics that have significantly impacted the United States, so I’ll assume you’re referring to one of the most infamous ones, the 1918 influenza pandemic, commonly known as the Spanish Flu.
The Spanish Flu of 1918 is estimated to have caused the death of about 675,000 people in the United States. However, these figures can vary depending on the source, as record-keeping wasn’t as precise back then, and the disease itself was not well understood.
If you were referring to a different influenza pandemic, please provide more details.
Although Artificial Intelligence (AI) is more than just chatbots—AI also underlies self-driving vehicles, drones, and robots in new niches—chatbots are driving most AI discussions lately. The articles I’ve read have barely mentioned the impact these chatbots could have on employment, but before we get to that, let me mention three chatbots I’ve used.
First, let me note available chatbots cannot perform any actions outside the chat box, such as making phone calls, sending emails, or ordering products. But as my son informed me, there’s nothing technological to stop a company from creating a chatbot that violates those constraints.
- ChatGPT from OpenAI sparked the current mania for chatbots when it was unveiled in November 2022. After some free usage, I subscribed for $20/month to guarantee access and further learn about chatbots and how to use them.
- ChatGPT 3.5 completed its training in March 2021 and knows nothing about events after that.
- It does not actively search the web for new information.
- OpenAI controls the frequency of your requests and has a usage agreement that bars use for political, spam, and other nefarious purposes.
- ChatGPT is like a coffee shop buddy who swore off news a few years ago, but keeps decided opinions which he shares forthrightly with you, often without specific references.
- Bing. Microsoft’s chatbot is built atop ChatGPT, to which many excellent features have been added.
- Bing continues to learn.
- It scours the web to augment its responses, which come with links and usually conclude with buttons to further information on the topic.
- Bing has three response modes: creative, balanced, and precise. I think of them as friendly artist, helpful librarian, and terse professor.
- If you select creative and request an image (e.g. show me a cyberpunk image of a computer typing on a keyboard). Bing will, using Image Creator—built atop DALL-E—also from OpenAI, create a visual response. See the robot at the top of this post.
- Bard. Google’s chatbot
- Bard has been independently developed by Google AI department. It is still learning.
- Bard rephrased a section of this post with as crisp an effect as had ChatGPT and Bing.
- When I asked about specific mutual funds, Bard returned an answer which misidentified the funds and gave incorrect historical information. I flagged Bard’s answer with the reasons why I gave its response a thumbs down.
- I ran a similar inquiry yesterday, which Bard handled well and properly.
- Bard does not yet have an image creation facility.
Likely Developments in the Next Twenty Years
Already we can see chatbots handling tasks that humans currently perform. For instance, first contact customer service. Chatbots will increasingly automate tasks that were formerly done by humans. Routine tasks like answering questions and describing steps to resolve issues are in a chatbot’s wheelhouse. This frees up human agents to focus on more complex tasks after retraining. Substituting chatbots or humans leads to significant cost savings and, often, improved customer satisfaction.
Other employment sectors, which will be affected, emerge when we consider where AI researchers are focusing their efforts.
- XAI. eXplainable AI is a field of research that seeks to develop AI systems that can explain their decisions in human terms. This is important in several domains, such as healthcare and finance, where AI systems are increasingly being used to make decisions that can have a significant impact on people’s lives. XAI systems can help to build trust between humans and AI systems. It can also help to identify and address potential biases in AI systems.
- Autonomous self-driving and drones are rapidly emerging technologies that have the potential to disrupt the transportation and logistics industries. These developments could lead to the loss of jobs for truck drivers, pilots, transportation and warehouse workers. However, they could also create new jobs in the development, maintenance, and operation of autonomous vehicles and drones. The net impact of these technologies will have a significant impact on the transportation and logistics industries.
In addition, there are a number of other ways that AI will impact the workforce. For example, AI is being used to automate tasks in manufacturing, agriculture, and other industries. AI is also being used to develop new products and services, which can create new jobs. The overall impact of AI will have a significant impact on the way we work and live in the future.
Compliance with AI regulations depends on the willingness of users, some of whom can gain by unrestrained use of the technology. Of course, devious humans directing chatbots to cheat or influence people to act to other people’s benefit remains possible.
Are chatbots, robots, and other AI creations magical assistants or devilish replacements?
According to a recent World Economic Forum report, employers expect to create 69 million new jobs by 2027 and eliminate 83 million positions because of automation and AI. That will cause a net loss of 14 million jobs in 5 years, equivalent to 2% of current worldwide employment.
Disruption in historic hiring practices in healthcare, transportation, warehousing, and financial services may also to be expected.
How will society adjust to the increased level of chronic unemployment?
- Educate children to be skilled in tasks that are not done well by AI or robots. Such tasks require personal contact and creativity. Scam detection and avoidance must also be added to the curriculum.
- Unemployment will coexist with greater societal wealth. The social welfare net will be tested. In my opinion, tax on corporate profits must be equal to or greater than that leveled on individuals.
- Humanity has experienced two epochal changes that vastly changed society’s structure.
- The Agricultural Revolution enabled settled communities and a surplus supporting a leisure class, which, at its most productive, created improvements to agriculture and lifestyle that eventually benefited all.
- The Industrial Revolution shifted the focus from farm to factory. Vast quantities of new goods became available as well as the surplus allowed increased numbers of people to work outside of production and develop new ideas, products, and services, which resulted in social changes.
Society has changed in response to industrial orientation. Not as quickly, not as thoroughly as many think it should have, but society has changed.
- Children are educated rather than placed in factories and mines.
- Workers no longer work 12 hours a day, 6 days a week, fifty-two weeks a year.
- Seniors are not expected to work until they died.
Will chatbots, AI, and robotic developments result in new societal changes? Perhaps a minimum guaranteed income for those unemployed who will never be able to gain work against mass-produced chatbots in factories run by robots.
Image of Robot typing on a computer. Generated by Bing from my prompt.
Future Epoch changes to societal surplus results in changes to society.
ChatGPT Try OpenAI’s chatbot.
Bard Try Google’s chatbot.