The term artificial intelligence can evoke feelings of wonder or apprehension, depending on who is discussing it.
No matter our feelings, it appears AI is here to stay, with the market forecast to reach $1,339 billion by 2030, as reported by MarketsandMarkets. A Forbes survey revealed that 77% of participants are concerned that AI could lead to job losses.
This concern is genuine, according to Manasi Vartak, the chief AI architect at Cloudera, but holding onto that fear could hinder workers; being open to new technologies may be a better way to secure employment.
What is generative AI?
Generative AI enables us to transform our thoughts into tangible outcomes, as described by Joe Edwards, the director of product marketing at UiPath, an automation software company.
This type of AI can produce text, music, images, or videos based on a few prompts. Its impact has been notable on social media this year due to the emergence of AI-generated art, faux celebrity images, and music created posthumously.
Generative AI operates by learning from vast datasets, observing images, videos, and online human behavior, according to Edwards. These sophisticated machine learning models recognize patterns and generate original content.
Previously, AI was primarily “predictive,” meaning it could assess the sentiment of a tweet or news article. Today, it can compose a complete news article, creating content from scratch, which marks a major shift, according to Vartak.
Generative AI has its benefits and drawbacks.
For instance, Vartak notes that it excels in writing. It can draft tweets, emails, or even elaborate fictional stories. It can simplify complex topics for summaries or assist with routine tasks like meeting transcriptions and data organization.
On the flip side, it sometimes becomes overly creative, leading to incorrect or misleading responses. For example, if you ask ChatGPT, “How many ‘r’s are in the word strawberry?” you might receive inaccurate answers consistently.
Edwards points out, “It may sound very sure of itself while it misleads you.”
When a large language model identifies nonexistent patterns or produces nonsensical responses, it is known as “hallucinating.” This presents a significant issue in this technology, Vartak explains. When asked to summarize an article, it might only capture about 80% accurately.
While some inaccuracies might be low-risk and amusing, others can propagate harmful misinformation. For instance, after Google launched its AI feature earlier this year, a widely circulated reply suggested that “Doctors recommend smoking 2-3 cigarettes a day during pregnancy.”
Moreover, since generative AI is created by humans, it can reflect human biases. A 2023 report from the Washington Post revealed that AI-generated images can reinforce damaging racial and gender stereotypes.
“These models are trained on existing data, which carries biases, so they might lack representation from women, minorities, LGBTQ communities, and people of color,” Vartak warns. “If we blindly accept what technology provides, we risk overlooking important narratives.”
How to utilize generative AI
Always involve a human in the process, Vartak advises. Generative AI should serve as a companion in your tasks, not take over completely.
“Provide it with ideas, let it generate some text, then review it for accuracy,” she recommends. “Trust, but verify.”
Edwards utilized a generative AI tool to sift through numerous emails after returning from paternity leave. It can also serve as a brainstorming ally for planning events like birthday parties or vacations.
Home cooks can use it to come up with new recipes, while musicians can try out custom music tracks. Some potential homebuyers are turning to it for assistance in their house-hunting endeavors.
There are also specific business applications tailored to various professions. For educators, this means honing skills to identify essays generated by ChatGPT among students. In the healthcare sector, doctors are leveraging AI to enhance patient consultations and translate clinical notes into understandable language. Certain AI systems even assist in detecting breast cancer, pulmonary embolisms, and strokes, potentially leading to earlier lifesaving diagnoses.
Companies like Cloudera and UiPath develop customized private AI systems that are trained on less extensive data to reduce risks like hallucinations and information leaks.
Both Vartak and Edwards encourage people to experiment with generative AI, whether it’s casually exploring ChatGPT or enrolling in an accessible online training course.
“Consider it as your trusty sidekick,” Vartak concludes. “It can enhance both enjoyment and productivity in your daily activities. While all new technologies can be challenging, this one is here to stay and has the potential to be beneficial.”
