No ratings yet.

Imagine a world where planning your next vacation is simple and straightforward, where an advanced system takes care of all the details for you, including finding hotels, flights, and activities, all based on what you like. Our team is excited to propose a new feature for Booking.com, a GenAI-powered travel planner that could completely change how people organize their trips and give Booking.com a real advantage in the online travel agency industry.

Presently, many travelers planning their next trip spend a lot of time looking through countless options, trying to pick the right one for them. This process can be overwhelming for some, and there can be times when great opportunities are missed out on. What we’re suggesting is a travel planner powered by GenAI that creates personalized itineraries and recommendations for users. It would consider their preferences, past bookings, and real-time data like local events or weather updates when providing suggestions, all in the effort to make trip planning easier and more enjoyable for users of the platform. Instead of wasting time on things they don’t care about, users would get suggestions that match exactly what they’re looking for.

Integrating this AI planner into Booking.com’s website and app would be seamless. Travelers access the travel planner via a blue icon found on the home screen of the Booking.com website or app, which pops up a chatbot interface that enables users to converse with the AI chatbot and inform them about their general requirements, such as where they’re going or what they’re interested in doing.

The AI behind the chatbot would then pull together relevant data from internal and external sources and immediately suggest the best options. From there the user can continue the interaction with the chatbot, similar to a travel agent, and develop a complete itinerary as per their needs. At any point if the user wants to make changes, they can just inform the AI chatbot for necessary modifications, and it will update the plan right away. The best part is that the planner can also adjust plans during the trip. If a flight gets delayed or the weather changes, the AI will notify the traveler and offer new suggestions, making sure everything goes smoothly. In the competitive online travel agency market, this ability to adapt in real-time would set Booking.com apart from competitors like Expedia, who are also exploring AI functionalities.

Travelers today want more than just a booking site, they want a tool that understands their needs and gives accurate and relevant recommendations. By integrating GenAI into what they offer, Booking.com would not be just another booking platform, but a complete travel planning assistant available at any time. For Booking.com, the benefits are huge. User engagement on the platform would grow, with more planners interacting with the AI and exploring recommendations. This could in turn lead to more bookings and unlock additional revenue streams from partnered services like car rentals or tours. The more users interact with the planner, the more appealing the platform becomes to service providers, who would want to list their services. As the AI improves its suggestions as a result of increasing options, customer satisfaction and loyalty will undoubtedly improve, which are both critical objectives for any business.

To put it simply, integrating this AI-powered travel planner would completely transform the way people plan their vacations. With AI transforming industries everywhere, now is the perfect time for Booking.com to take travel planning to the next level.

Team 22: Emma Montag, Liam Albachre, Olivier Leurink, Amit Abraham

No ratings yet.

Walking into a gym for the first time can feel daunting, with several intimidating machines, weights, and fitness equipment that can leave you a little overwhelmed at the start . As someone who had never gone to the gym until the summer of last year, I was quite lost on what exercises I should be doing or even how to use half of the machines I saw when I first entered. With ChatGPT getting a lot of traction last year, I thought of running a casual experiment just to see how helpful it could be as a ‘virtual’ gym trainer, since I definitely wasn’t going to pay for an actual one, and it really helped a lot more than I expected.

My initial interactions with ChatGPT were to help me build a simple workout plan based on my fitness level and the gym equipment that was available at the time. Being pretty much a beginner in this, I wasn’t looking for anything too complicated, just something that would break down the basics for me without making me feel lost. GPT was great with this, suggesting easy to follow routines that introduced me to the most common gym machines and free weight workouts. It would also explain the proper form when using such machines and the muscle groups they targeted, giving me a better understanding of what, why, and how I should be doing things.

What sets it apart from a Google search that usually gives generic routines or advice, is the ability to constantly refine things based on the information I gave it about myself. I would let it know things about myself like for example, my goals and how much time I had on days, which it would adapt to and set plans based on this. There were some days when I barely had the time, where I’d tell it that I have max 30 minutes for a session, and it would modify the routine taking into account the workouts I have already done in the week and what exercises to prioritize that day. It literally felt like I had a personal trainer on standby.

However, unlike an actual trainer, ChatGPT can’t track my progress, like how much weight I would be lifting or how well I’m performing each set. If it had a way to keep tabs on my progress without me needing to input such kind of information all the time, it could be a gamechanger in this space. This is why I feel the camera viewing ‘vision’ capability that OpenAI has shown in their tech demos but not officially released yet (OpenAI, 2024), could really make an impact here. Right now the AI model is limited to analyzing uploaded images, but the potential to integrate live camera feeds or at least video uploads would be huge. I can imagine pointing the phone camera to show your form as you perform each rep and have it analyze and offer corrections or track form improvements over time. This would literally make it an interactive, intelligent trainer and not just a source of workout advice.

As it stands, ChatGPT gave me what I was looking for in terms of a plan, guidance around things, and the flexibility to make adjustments for me, but it’s not hard to think about where it could go next. The potential is clear, and with some added capabilities, all I can say is that personal trainers and coaches need to watch out.

References

OpenAI. (2024, May 13). Live demo of GPT-4o’s vision capabilities [Video]. YouTube. https://www.youtube.com/watch?v=rKp36MmRlXA

5/5 (1)

Artificial intelligence has made its footing in the stock market, driven by algorithmic trading at its very core. More and more investors and hedge funds now turn to the might of AI-driven algorithms for quicker, more incisive decisions on trading. These algorithms can process huge quantities of information, identify patterns, and execute trades in microseconds, something no human can do. The question that piqued my interest was whether these algorithmic models can outperform even today’s best traders.

In many of these cases, the answer to the question would be an easy yes, as seen in the case of the hedge fund Renaissance Technologies, popularly known for its Medallion Fund, which makes investment decisions based on AI-based quantitative models. The fund has been able to maintain better performance than its human counterparts, with an average annual return of approximately 66% over the last two decades (S., 2024). Another example is Bridgewater Associates, which uses AI to critically analyze global economic data to assist in trading (Jensen & Moriarty, 2024).

The real edge that AI has over humans is the fact that it can analyze both structured and unstructured data on the stock market, including in the news, social media, and financial reports, enabling more accurate predictions for future market movements (Gnibus, 2024). Furthermore, these insight-driven trades are executed much faster than any human trader. This saves a great amount of time, and as they say in the world of trading, time is money. With great speed and precision, AI-driven trading has now taken Wall Street by storm, with 60% of all trades being executed by algorithms (Schlaepfer, 2024).

However, AI isn’t foolproof per se. The most critical issue it faces is the dependence on historical data. Although this may sound normal, the over-reliance on past information does not help in the case of unprecedented events like the 2008 financial crisis or the COVID-19 pandemic, leading to poor decisions in such scenarios. Programs trained on old data can often fail to adapt to the abrupt market swings and eventually cause large losses. Additionally, AI-controlled strategies contribute to volatility, as seen in the case of a “flash crash”, when automated high-frequency trades create huge, unexpected market fluctuations, and can lead to major setbacks for many players (Needhi, 2024).

I personally feel that although AI has tremendous potential, it still needs to be used with and not instead of human traders. Rather than replacing human counterparts, AI should be used to enhance stock market decisions ultimately made by humans. What are your thoughts on this – will AI become the norm when making trading decisions and eventually dominate the stock market?

References

Gnibus, A. (2024, April 15). Why Unstructured Data is Having a Moment, Thanks to AI. Medium. https://medium.com/@alexgnibus/why-unstructured-data-is-having-a-moment-thanks-to-ai-ad546e5a7158

Jensen, G., & Moriarty, J. (2024, May 30). Are we on the brink of an AI investment arms race? Bridgewater. https://www.bridgewater.com/research-and-insights/are-we-on-the-brink-of-an-ai-investment-arms-race

Needhi, J. (2024, July 9). AI’s Role in the 2024 Stock Market Flash Crash: A Case Study. Medium. https://medium.com/@jeyadev_needhi/ais-role-in-the-2024-stock-market-flash-crash-a-case-study-55d70289ad50

S, R. (2024, September 27). How Jim Simons built the best AI that solved the market. Medium. https://suenrankine.medium.com/how-jim-simons-built-the-best-ai-that-solved-the-market-f3e23319db29

Schlaepfer, D. (2024, July 2). Humans vs. AI in The Stock Market: The Worst Trade Ever Made? Forbes. https://www.forbes.com/sites/forbesbooksauthors/2024/06/18/humans-vs-ai-in-the-stock-market-the-worst-trade-ever-made/