Hugo Pellegrin

🎭 Convious AI, a chatbot by Convious, aims to increase online ticket conversion rates for amusement parks by integrating with web messengers like WhatsApp. Offering 24/7 support, features include ticket purchasing, modifications, instant reissues of lost tickets, and other customer assistance without operational effort. Supporting all languages, the tool caters to international tourists and can generate additional sales like restaurant bookings and rentals. CEO Camiel Kraan highlights its ability to personalize and streamline the ticket buying process, enhancing the visitor experience. Currently in beta, Convious AI will be showcased at IAAPA Europe in Amsterdam at the end of September.

💻 The National Museum of Computing at Bletchley Park, UK's main WWII decryption site, is reviving Alan Turing using AI, a controversial project. Visitors to the museum can converse with an Alan Turing avatar, developed by 1956 Individuals in collaboration with video game industry artists. Critics, including New Scientist, label the project "imprudent" and "problematic". The AI-powered exhibition, Never Alone, awaits launch in a few months.

📈 Over 80% of generative AI projects fail, doubling the average failure rate for tech projects, according to a Rand Corporation study interviewing 65 data scientists. The study identifies five main failure causes and offers best practices. Key failures include misunderstanding the problem, lack of data for training models, and inadequate infrastructure. Success requires clear problem definition, proper data and tools, and long-term commitment. Understanding AI's limitations is crucial for project leaders.