AI can help prevent gun violence. Here's how ...

Good morning, lovers of french fries and AI news! Let’s jump into today’s stories, which outline inspiring AI developments. 🗞

🔨 AI TOOLS OF THE DAY

🎨 ArtPrint: Create unique artwork for your home by describing your desired art style. → check it out

🗣 Binko: An advanced language translation and chat app. → check it out

🖥 Legit-startups: An AI-powered tool that generates personalized startup websites, tailored to any topic or idea, ready for launch. → check it out

CAN AI HELP PREVENT GUN VIOLENCE? 🙏

Major video recognition innovations in AI have the potential to save hundreds of lives. 🙏

What’s happening? Miami’s Frost Museum of Science is using AI to detect guns and identify possible shooters as quickly as possible to thwart potential mass shootings.

How does it work? The museum has programmed AI-powered visual recognition features into their existing cameras to spot different types of weapons. If a potential weapon is detected above a certain threshold of suspicion, the entire security staff is immediately notified and the exact location of the individual is pinpointed and tracked.

Can this actually help? The VP for technology at the museum stated, “Every tool helps, every second counts. We will do anything that we can do to further protect the community and our visitors and staff.” The president of the company behind the tech said this could be incredibly useful in places like airports, shopping malls, schools, and other crowded spaces with limited security. He said, “You cannot have millions of people watching cameras all over the world—it’s almost impossible.” The goal is to get responders and law enforcement officers to the scene as quickly as possible to prevent anyone from being harmed.

MACHINE LEARNING? WHAT ABOUT MACHINE UNLEARNING? 🤔

We need to make AI algorithms dumber by making them smarter … or smarter by making them dumber … something like that. 🤯

What’s happening? Google has announced a machine unlearning competition to take place from mid-July to mid-August. The competition will be hosted virtually on Kaggle, and submissions will be automatically scored “in terms of both forgetting quality and model utility.” Google hopes that this competition will help spark innovation in machine unlearning and encourage the development of ethical unlearning algorithms.

What is machine unlearning? Machine unlearning is an emergent subfield of machine learning that aims to remove the influence of a specific subset from a trained model. Machine unlearning is aimed at protecting user privacy, erasing inaccurate or outdated information from trained models, and removing harmful, manipulated, or outlier data.

Why is Google doing this? Google seems to be responding to the immense public pressure displayed by experts who warn of the risks of AI towards data misuse and cybercrime. They are also looking to comply with incoming regulations proposed by the European Union which are aimed at protecting privacy.

Why could this be useful? Machine unlearning aims to give individuals more control over personal data and allow users to wipe out their information from a given algorithm, protecting them from privacy threats and keeping others from misusing their data.

AI IS TRANSFORMING THE WAY WE RECYCLE ♻️

A lot of talk has been about how AI is going to end the world, but what if AI could actually be used to save it? 🌏

What’s new? GreyParrot, a UK start-up, has created an AI system designed to analyze the content being filtered through waste processing and recycling facilities.

How does it work? GreyParrot placed their cameras above the conveyor belts of around 50 waste and recycling sites in Europe, utilizing AI software to analyze what passes through in real-time.

What is the purpose? Approximately 2.24 billion tons of solid waste was produced in 2020, and this is on rate to increase by 73% to 3.88 billion tons by 2050. GreyParrot's systems are projected to track 32 billion waste objects each year, and the firm has built up a huge digital map of waste. This data can be used by managers of waste facilities to become more efficient in their operations, but the main intent is to use the data to inform regulators and producers of which packaging is most harmful and which is most useful in recycling efforts. Mikela Druckman, the founder of GreyParrot, said, "It is allowing regulators to have a much better understanding of what's happening with the material, what materials are problematic, and it is also influencing packaging design.” Her hope is that large brands and other producers will start using data generated by firms like GreyParrot to learn how to design more reusable and environmentally friendly products.

HAS AI TAKEN OVER THE WORLD? CHECK OUT THE FRY METER BELOW