Binit's founder, Borut Grgic, claims that the device's trash recognition is almost 98% accurate after integrating OpenAI's image recognition capabilities. The startup also plans to offer premium features via subscription and position itself as a data provider on waste habits. The device is currently being piloted in five US cities and four European cities, with a targeted price point of around $199.
Key takeaways:
- Finnish startup Binit is developing an AI gadget that uses large language models' image processing capabilities to track household trash, aiming to boost recycling efficiency.
- The AI hardware, which is designed to be placed in the kitchen, uses cameras and sensors to scan items before they are thrown away, and uses OpenAI’s GPT for image recognition.
- Binit's system provides analytics, feedback and gamification via an app to encourage users to reduce their waste. The startup also plans to sell data on people's waste habits to entities like packaging companies.
- The startup has been testing its AI trash scanner in nine cities across the US and Europe, and is planning a commercial launch in the US this fall, with the hardware priced at around $199.