Navigating AI Frontiers: Scarlett Johansson, Travel, and Beyond

AI development has seen remarkable progress in the past year, with advancements in machine learning, natural language processing, and computer vision driving innovation across various industries. The widespread adoption of AI technologies is going to transform traditional business models, enhancing productivity, enabling personalized user experiences, and change how we work. However, we also need to ensure there is mechanism to address potential societal challenges while harnessing its full potential.

Artificial Intelligence vs Intellectual Property

You might have heard about the newest OpenAI model, dubbed GPT-4o, which was released ten days ago on May 13. Not only does it work faster than previous versions and reason across text, audio, and video in real time, but it can also assess a person’s emotional state by analyzing a selfie video of their face. However, the release was somewhat overshadowed by the fact that one of the AI personal assistant's voices was uncannily similar to Scarlett Johansson's. What was more shocking was that Johansson stated she had declined a request from OpenAI nine months ago to license her voice for the new ChatGPT voice assistant. Now, a lawyer is involved, and OpenAI has announced that they have taken the assistant offline.

Johansson seems to have a case here, but the intersection of artificial intelligence (AI) and intellectual property (IP) presents many complex challenges. From copyright attribution to patentability and data ownership, the regulatory framework governing AI within IP is not keeping up with the emerging and fast-evolving issues necessary to safeguard the rights of creators, innovators, and users. Most companies do not have any clear strategy or policy on using artificial intelligence.

When social media became popular a decade or so ago, many large companies had very strict policies against employees using social media because of potential risks for information leaks and legal exposure. Many companies have since reversed their stance and now encourage employees to use social media to promote company business. The difference is that AI evolves at a much faster pace. A survey published this month by LinkedIn suggested that three-quarters of white-collar workers have used AI for work, and four-fifths of them have done so from their own accounts and devices. I suspect many employees have already been using ChatGPT for reports, emails, memos, etc. When I visited China, I saw people using similar AI tools for work plans and consulting reports. Most companies probably do not yet have a good understanding of the long-term implications.

AL Will Boost Productivity

Meanwhile, I tried ChatGPT for travel planning, and it worked magically well in terms of efficiency. Here is a brief description of the steps I used for the travel plan, which would otherwise have taken me a few weeks to devise.

Me: Can you propose an itinerary for a ten-day self-driving trip in Northern France?

ChatGPT provided a day-by-day plan from Paris to Rouen, Honfleur, Bayeux, the D-Day Beaches, Mont Saint-Michel, Saint-Malo, Dinan, Caen, Giverny, and back to Paris. It included the exact distance between each stop and the driving time.

Me: Can you add some specific suggestions for a photographer onto this trip?

The output was not comprehensive, but it suggested some interesting photography spots for each day, some even including the specific lens to use.

Me: How about including some reasonably priced places for dinner and lunch, and museums/art galleries?

It quickly added a list of restaurants for each day, including the type of food and specialties, and potential activities for each day.

A few days later,

Me: Can I continue the conversation about the France self-driving trip itinerary? I would like to include Burgundy in the trip, still a ten-day overall trip. I also would like to add Maison Picassiette to the itinerary.

ChatGPT produced a detailed revised itinerary in seconds. With a total of less than 30 minutes of work, ChatGPT generated an excellent starting point for planning the trip. It would have taken me days, if not weeks, to read through tour books to gather and sort through similar information.

AI Model Is Inherently Biased Due To Human Biases

While the trip planning exercise was very efficient, I quickly realized that all the suggested options, including the photography ideas, tended to be popular tourist places. This is a natural result of how the AI model was trained and how the algorithm is designed to work. If you are trying to find unique experiences to fully immerse in the culture, you have to be cautious about how you use AI-generated information.

In a more general sense, it is only natural that human biases have made their way into AI systems. When biases are embedded in the model, the rapid development of AI will only amplify the resulting effects. As AI is expected to be quickly deployed into areas such as people development and organizational issues for productivity, avoiding the unintentional perpetuation of human bias based on historical training data will be a challenge for organizations to address.

In summary, AI is indeed quickly changing how we work and at an unthinkable pace. The revolution will reshape the world. Let's all hope for the better.

Related Articles

There is a Ditch on Both Sides of the Road (1/10/2019)

Before You Innovate, Ask Questions (12/05/2020)

(Please join the 349,000 subscribers to my newsletter by clicking the?subscribe?button to be notified of new editions. Better yet, please join the conversation by sharing your comments. Please continue to forward it to anyone else who might enjoy it. You can also provide your support by purchasing my book: The Odyssey of Self-Discovery: On Becoming A Leader .)