AI and You Hallucinations, Big Tech Talk on AI, Jobs

AI and You Hallucinations, Big Tech Talk on AI, Jobs

The Quirky and Imaginative World of Generative AI

Exploring the Strange, Creative, and Mischievous Side of AI

Over the past few months, I’ve immersed myself in the fascinating realm of generative AI, delving deep into AI glossaries to uncover the vocabulary associated with this new world. As an expert in technology and media, I must admit that I may know more about AI than the average American. However, a recent conversation with a journalism professor made me realize that some key concepts related to generative AI are not widely known or understood. When I mentioned that AI tools have a tendency to “hallucinate,” they stared at me blankly, asking, “What does that mean?”

To put it simply, “hallucinate” is a term that everyone should be familiar with in the context of generative AI. It refers to the phenomenon where AI engines, such as OpenAI’s ChatGPT, generate content that sounds true but is actually fabricated. In fact, OpenAI recently faced an investigation by the US Federal Trade Commission for potentially making false statements through its chatbot. The term “hallucinate” was coined by Google DeepMind researchers in 2018, who found that neural machine translation systems can produce translations that are completely detached from the source material.

Although glossaries provide more straightforward explanations of AI “hallucinations,” it’s worth noting the definition offered by NBC News: “Hallucinations are unexpected and incorrect responses from AI programs that can arise for reasons that are not yet fully known.” These hallucinations can range from providing fruit salad recipes instead of information on planting fruit trees to fabricating scholarly citations and making up facts about events.

To gain a better understanding of how AI perceives “hallucinations,” I decided to ask ChatGPT for its input on the matter. In its verbose response, ChatGPT described “hallucinate” as the generation of content or information not based on direct observation or real-world data. The term was borrowed from the field of psychology, where hallucination refers to sensory experiences that occur without external stimuli. In the AI context, it means producing outputs that deviate from reality, introducing elements not present in the original data. While ChatGPT’s response might seem long-winded and defensive, it highlights AI’s potential to go beyond what’s directly observed or instructed, resulting in imaginative and unexpected outcomes.

Now that we’ve explored the fascinating world of AI hallucinations, let’s delve into other noteworthy developments in the field of AI.

US Workers’ Attitude Towards AI and Potential Job Erosion

The Pew Research Center and McKinsey released reports this week shedding light on how AI may impact workers and jobs in the US. According to Pew, US workers generally have a positive outlook on AI’s impact on their jobs. Many workers in industries exposed to AI believe it will benefit them personally, outweighing concerns of job loss.

Pew categorizes jobs as more exposed to AI if AI can either fully perform their main activities or aid in them. Job loss due to AI remains uncertain since it can both replace and complement human work. Managers ultimately decide how and when AI tools are employed, making decisions based on the potential to enhance productivity or create new job opportunities.

Industries with the highest exposure to AI align with expectations, but it may surprise some that women are more likely to have exposure to AI in their jobs. McKinsey’s report, “Generative AI and the Future of Work in America,” predicts up to 30% of current work hours in the US could be automated by 2030, with potential occupational transitions required.

An estimated 12 million occupational transitions may be necessary, and lower-wage workers face a higher likelihood of needing to change occupations. Women are also 1.5 times more likely than men to require a transition. These transitions are dependent on employers helping to train workers and extending opportunities to marginalized groups.

In the context of AI-driven job recruitment, companies are employing AI tools to aid in hiring decisions. If you’re interested in navigating the job market in this AI era, I recommend checking out ENBLE’s Laura Michelle Davis’ guide, “The New Age of Hiring.”

AI Takes Center Stage in Big Tech Earnings Calls

During their quarterly earnings calls, major tech companies including Google/Alphabet, Microsoft, and Meta (formerly Facebook) prominently featured discussions on AI in their opening remarks and analyst Q&A sessions. These tech giants recognize the transformative power of AI and its relevance in their respective industries.

Microsoft CEO Satya Nadella acknowledged AI’s significance, mentioning it 27 times. Sundar Pichai, Google’s CEO, emphasized the role of AI in revolutionizing Google Search and mentioned it 35 times. Mark Zuckerberg, CEO of Meta, highlighted AI’s growing presence in Facebook’s content recommendations and numerous potential applications across their platforms.

These companies are leveraging AI to enhance search experiences, connect people, and provide creative tools that facilitate content sharing and interaction. AI’s ability to connect ideas and preferences holds promise for personalized recommendations, decision-making support, and deeper exploration of topics on the web.

AI Detection Faces Challenges and Promises

In a previous article, I discussed Stanford University researchers’ attempts to develop AI “detectors” capable of distinguishing content created by AI from human-generated content. Unfortunately, their efforts proved fruitless, highlighting the biases and vulnerability of the software.

In January, OpenAI, the creator of ChatGPT, announced its development of an AI detection tool. However, Decrypt reported that OpenAI quietly unplugged the tool due to its low accuracy. OpenAI stated its commitment to researching more effective techniques and eventually providing users with mechanisms to identify AI-generated audio or visual content.

While progress is being made, AI detection still faces challenges in ensuring accurate identification. The development of robust and reliable tools is necessary to build trust in AI-generated content.

AI Regulation Discussed by US Government

Senate Majority Leader Chuck Schumer continues to lead discussions on AI regulations, advocating for legislation that fosters innovation while mitigating potential risks. Bipartisan interest in shaping AI legislation indicates a shared recognition of the importance of safeguards to prevent liabilities associated with AI applications.

As the Senate plans to call in experts for further testimony, concerns about AI being used for malicious purposes, such as biological attacks, have been voiced by senators on both sides. It’s worth noting that science-fiction movies have previously explored such scenarios, reminding us of the need for responsible AI governance.

Hollywood Marches On Despite Strikes, AI Talent in Demand

Against the backdrop of ongoing strikes by actors and writers, entertainment companies are seeking AI specialists to fill roles traditionally occupied by creatives. The concern among professionals in the entertainment industry lies in studios using AI to reuse their likeness or voices without proper compensation or acknowledgment.

Nearly every major studio is embracing AI for diverse applications, such as content creation, customer service, and data analysis. Disney, Netflix, and Amazon are actively seeking AI talent for various positions, with salaries reflecting the importance of pushing the boundaries of AI capabilities.

The implementation of AI in the entertainment industry comes with the inevitable prospect of job displacement and questions regarding fair compensation for all contributors, including actors and writers, among others. Actor Joseph-Gordon Levitt highlighted the importance of recognizing and rewarding those whose work serves as the foundation for training AI models.

These developments within the entertainment industry and ongoing discussions surrounding AI regulation demonstrate that the decisions made by humans will shape the future of AI.

As we navigate a future intimately intertwined with AI, it becomes increasingly essential to comprehend its quirks, challenges, and potential. The imaginative world of generative AI presents us with captivating possibilities and unexpected situations, pushing us to explore the limits of our creativity and responsibility.