Intrigued by the capabilities of A.I.-powered chatbots like ChatGPT and Bard? These chatbots can impressively write essays and recipes. However, there is a common issue with artificial intelligence known as hallucinations, where it tends to fabricate information. These chatbots, which rely on information gathered from various sources on the internet, often make errors. Publishing a cake recipe with incorrect flour measurements is an example of this, and it can be disappointing.
As A.I. continues to integrate into mainstream tech tools, it is crucial to understand how to utilize it effectively. After testing numerous A.I. products, I have discovered that most people are using the technology in suboptimal ways, largely due to inadequate guidance from tech companies.
The chatbots are least helpful when we ask them questions and expect accurate answers based on their own reasoning, as that is not how they were designed to be used. However, when we direct them to use information from trusted sources like credible websites and research papers, A.I. can perform useful tasks with a high level of accuracy.
Sam Heutmaker, the founder of Context, an A.I. start-up, explains that chatbots can do interesting things when provided with accurate information. On their own, approximately 70% of what they produce is likely to be inaccurate. By advising chatbots to work with specific data, we can elicit intelligible answers and valuable advice.
I experienced a transformation from being a skeptical A.I. user to an enthusiastic one when I used a travel itinerary planned by ChatGPT. The recommendations in the itinerary were sourced from my favorite travel websites, which resulted in a successful trip.
Directing chatbots to specific high-quality sources, such as well-established media outlets and academic publications, can also help combat the spread of misinformation. I will share some of the approaches I used to seek assistance with cooking, research, and travel planning.
When it comes to meal planning, chatbots like ChatGPT and Bard may generate visually appealing recipes that fail in practice. However, I discovered that using ChatGPT plug-ins, third-party apps that work with the chatbot, can lead to better results. By selecting plug-ins like Tasty Recipes, which pull data from trusted websites like BuzzFeed’s Tasty, I obtained inspiring meal plans with accurate recipes.
For recipes from other publications, I used a plug-in called Link Reader, which allowed me to paste web links and generate meal plans using recipes from credible sites like Serious Eats. Although it required some extra effort, this method was preferable to relying on A.I.-concocted meal plans.
When conducting research on a popular video game series, I initially turned to ChatGPT and Bard for assistance in summarizing the plots. However, they made mistakes regarding important details. After testing various A.I. tools, I concluded that focusing on trusted sources and double-checking the data for accuracy was crucial. I found an effective tool called Humata.AI, a free web app popular among academic researchers and lawyers. It allows users to upload documents such as PDFs, and a chatbot then answers questions about the material while highlighting relevant portions from the document.
To test the tool’s capabilities, I uploaded a research paper from PubMed, a government-run search engine for scientific literature. In just a few minutes, the tool provided a relevant summary of the lengthy document, saving me hours of work. I could quickly verify the accuracy of the summaries by glancing at the highlights. Cyrus Khajvandi, the founder of Humata, developed the app when he was a researcher at Stanford and needed help comprehending dense scientific articles. He emphasized that chatbots like ChatGPT rely on outdated models of the web, which can result in a lack of relevant context.
When a travel writer from The New York Times asked ChatGPT to compose a travel itinerary for Milan, the bot made mistakes and led her to deserted areas. However, I had success when requesting a vacation itinerary for myself, my wife, and our dogs in Mendocino County, Calif. By incorporating suggestions from trusted travel sites like Thrillist and The Times’s travel section, ChatGPT generated a tailored itinerary that included dog-friendly options. This saved me several hours of planning, and most importantly, our dogs had a fantastic time.
In conclusion, companies like Google and OpenAI are working on reducing hallucinations in their chatbots, but we can already benefit from A.I. by taking control of the data they rely on for answers. Training machines with extensive data sets allows them to simulate human reasoning through language. However, it is vital to pair this ability with high-quality information, as emphasized by venture capitalist Nathan Benaich, who invests in A.I. companies.
Denial of responsibility! VigourTimes is an automatic aggregator of Global media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, and all materials to their authors. For any complaint, please reach us at – [email protected]. We will take necessary action within 24 hours.
