Most of the breathless hype you hear about artificial intelligence (AI) comes from those who have just used ChatGPT for the first time. ChaptGPT is a large language model developed by OpenAI that produces very plausible textual responses to open-ended queries. Its responses are so believable that many are predicting that the technology, which is continually improving, will replace up to 80% of human jobs within two years! The end of humanity, or at least of unoriginal writers, it seems.
To counter some of the hype, and to reassure readers who most likely are currently employed — but within the 80% predicted to be unemployed within two years — here are half-a-dozen myths relating to AI.
1. AI is the same thing as ChatGPT.
Artificial intelligence is a very broad term that refers to systems that display intelligent behavior by analyzing their environment and taking actions — with some degree of autonomy — to achieve a specific goal. In this context, "intelligence" is not the same thing as "sentient" or "human." Rather, it means "learning" from environment.
ChaptGPT is one of many AI applications. AI has been used for years in other contexts — for example, developing insights from large data sets and even social media. Many companies have for years relied on AI technologies to optimize processes and even invent new products. IBM's Watson comes to mind.
2. AI will destroy the world.
Perhaps AI will one day run killer drones (like in Terminator), but for now, it is another technological tool that has numerous beneficial applications. Nevertheless, the prospect of autonomous killing machines had the UN Convention on Certain Conventional Weapons sufficiently alarmed as to recommend banning the technology in weapons systems.
3. AI will take away all jobs.
The adoption of AI will undoubtedly result in the replacement of certain jobs, but it is likely to create others. I spoke with one PhD roboticist and asked her which jobs she thought would be lost and which would survive, and she answered surprisingly that jobs like plumbing would likely survive the longest, as it is technically very hard to train a robot to fix a leaky pipe.
Incidentally, self-driving appears to be farther off than we may have initially believed. The technical challenges and the legal and moral frameworks (who pays if an AI system crashes a car?) are proving to be difficult.
4. AI means that all of us will become lazy and have AI do our work for us.
Despite the hype, current AI products such as ChatGPT do not replace the thought and care needed to produce excellent writing. The output tends to be cliched and relatively easy to spot. Plus, it tends to invent citations and references and is frequently inaccurate. It relies on statistical models of probability to create what "sounds good" to human ears, and apparently that sometimes includes making up stuff.
5. AI has become "alive" and pulling the plug is like killing a living being.
Our existing tests for "humanness" are proving inadequate for the task of determining when our AI technologies have become "sentient." Ethicists, theologists and philosophers are asking these questions, which involve intangible qualities such as empathy, morality, and belief. The so-called Turing Test is probably not up to the task. Humans sometimes do dumb, or counterintuitive things. So do machines, apparently. One prominent ethicist and philosopher recently said that the concept of artificial "humanity" seems to involve the machine developing a sense of "morality," which is extremely hard to pin down through any one test.
6. Your company needs an AI policy.
This is true. Every company should be thinking about how AI is being used by employees, vendors and customers, and how to nurture beneficial uses and prevent substantial risks while avoiding overly restrictive policies that hold back beneficial uses and technology development. The best policies will be contained in a short, living document, that is frequently revisited and revised to keep up with the latest technologies and uses. To develop a policy, you need to understand how AI is currently being used within your organization and what the hopes for such uses might be in the future.
Don't worry that you are behind if you do not already have a policy. Many businesses do not. There are not many decent forms floating around that can be adapted. But it is a good time to create one.
I choose to be optimistic rather than alarmist about AI. Every new technology brings risks and opportunities.
Twenty years ago, we predicted Google Search would erode brains. Twenty years before that, the pocket calculator was blamed for the demise of students' math skills. What we later found, however, was that these remarkable tools enabled new insights and ways of doing things. In short, humans adapted and recalibrated to the capabilities these new tools and technologies enabled. Had we banned their use, imagine where we would be today.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.