Artificial intelligence dramatically filtered into the public consciousness a little more than a year ago, when OpenAI’s GPT-3, connected to its ChatGPT large-language model, began allowing the average person to type in questions and get comprehensive answers at lightning speed. Since then the rapid development of AI systems by all the leading tech companies has reinforced the power of this technology. Now it’s poised to affect everyone, far beyond the chatbots and background computer interactions going on for several years.
As the wife of an AI developer and someone who has been tracking AI advances for seven years, I can attest to the breathtaking revolution underway since the release of GPT-4 and now its latest version, one that demonstrates generative AI “thinking,” which OpenAI hasn’t even publicly acknowledged.
The speedy advances in AI development have inspired two schools of thought for the future: the accelerators (referred to in the industry as “e/acc” for “effective accelerationism”) and the decelerators (“e/a” for “effective altruism”). AI influencers in several high-profile companies, including OpenAI, and researchers at prestigious universities with AI programs signed an open letter last March advocating for a six-month moratorium on AI development till the ethical and moral issues are resolved. That went nowhere, for several reasons.
First, no AI company wants to suspend progress, especially since they all have investors expecting financially beneficial breakthroughs.
Second, if US companies were the only ones to suspend development, what would happen in places like China, not subject to a moratorium? They could pass the US in development, which our government does not support. Regardless of public opinion, our military will push forward with AI applications for weapons and other systems.
Right now AI is operating systems in the US that are mostly below the radar. An article in The New Yorker sums it up succinctly: (AI systems masquerading as people) “offer us customer service on Web sites, target us in video games, and fill our social-media feeds; they trade stocks and, with the help of systems such as OpenAI’s ChatGPT, can write essays, articles, and e-mails.” In addition, AI systems are already replacing jobs, though most are at a level that isn’t impacting the economy — yet.
In results of a survey by hiring website Resume‐Builder.com, 750 business leaders said they had replaced 37% of workers in 2023 and 44% said they plan layoffs in 2024 as AI replaces more employees. While it’s not a scientifically rigorous study, the results indicate that change is coming to our workplaces, something that’s been projected by the US Department of Education and Bureau of Labor Statistics. The BLS cites the jobs most susceptible to disruption by early AI as personal financial advisers, interpreters and translators, surgeons, fast-food workers, janitors, landscapers, truck drivers, and freight and stock laborers. Last year, too, Hollywood screenwriters and actors included protections from AI job-replacement in their WGA and SAG-AFTRA union agreements because they understand that AI threatens to wholly imitate their work and replace them, as both writers and digital characters.
On the other hand, workers are benefiting from the use of AI systems in several fields. Programmers are already finding that GPT can code far faster than they can, often flawlessly. However, the systems still need correction and guidance. In addition, while marketing companies find that chatbots can spit out proposals extremely quickly, they still need humans to review and reject those that are just too, well, robotic for their purposes.
But the longer these systems are trained, either in the public realm or by individuals who customize them for their company’s use, the more advanced they become and the better their skills for doing specific jobs. Recently OpenAI released GPTs, which enables people with no coding experience to create their own custom chatbots for applications like building a creative-writing coach for their own work, advising them on marketing plans, or any number of practical workplace uses and strategies. These tools typically cost $20 per month.
An enterprise version of ChatGPT became available in August 2023 for corporations, giving them the ability to create their own private, rapid systems customized for their workers. Many other large-language models are now in use, including dozens of customer-service chatbots and enterprise systems from Amazon, Alphabet (Google), Meta (Facebook) and Apple.
These will be game-changers in evolving or replacing jobs.
It’s only a matter of time before everyone will find AI useful, primarily because it will be as easy to use as Siri or Alexa. Companies like Inflection.ai, with its personal AI assistant called Pi, already employs a fluid, conversational mode of voice interaction as its primary interface. Whether that idea sits well with the public or not is irrelevant, just as someone’s love or hate for smartphones is irrelevant. AI will be just as ubiquitous.
