🤥 AIs are pretty good liars

🔮 Purple Horizons Dispatch

Hey there, Explorers!

What happens when AIs lie to us? For some people, like students, it can lead to getting an “F” or facing disciplinary action. For others, it can cost them their jobs. Like Thomas Grant Neusom, a Florida lawyer who used AI and was suspended from practice after filing cases that were “completely fabricated.” Given these potential consequences, can we truly trust AIs?

Today’s issue explores why AIs lie and what can be done to prevent this. We also look into AI doctors, why AIs think you want bacon ice cream, and how robots are learning from you.

Here’s what’s going on…

🗞️ TECH & TRENDS | News & More

🤥 AIs are pretty good liars | Sometimes AIs straight up lie to us. Not little white lies, either. Like completely made-up sources, non-existent people, or fake laws kind of lies. But why? The technical term is “hallucinations,” but according to MIT, we only call it that when we notice it’s wrong. The problem is that large language models are so good at what they do that what they make up looks right most of the time. And that makes trusting them hard. But they are getting better every day.

👩‍⚕️ Would you choose an app or a doctor? | If a doctor and an AI app both have 95% accuracy, what’s the difference? Why wait 10 days to see a dermatologist when an AI app with nearly the same accuracy as a clinician can treat you the same day? That's a question providers may soon have to answer.

🥓 Do you want bacon ice cream with your fries? | Most drive-thru experiences are less than optimal. This is one problem AI can surely fix. Or can it? Wendy’s and White Castle have had promising results, citing improved efficiency, cost savings, and customer satisfaction. Meanwhile, McDonald’s just abandoned their AI drive-thru experiment with IBM, removing the AI systems from 100 test locations after dealing with multiple (kinda hilarious!) order errors. Maybe they can just use AI to tell us if the ice cream machine is working or not? 🤷🏼‍♂️ 🍦

👀 Oh, oh… the robots are watching | Researchers from Stanford have developed a method for training humanoid robots to perform complex tasks like playing piano, tennis, and boxing by observing human movements. And the more they watch you, the better they get, too! This can potentially speed up development and reduce costs for future bi-pedal robots from companies like Figure and Tesla.

💡🤖 AI-Powered Success Starts Here | Thinking about AI but need a game plan? Looking for innovative ideas for your business? Purple Horizons is the innovation lab that helps you turn potential into profit. Book a 30-minute strategy session with Ralph & Gianni and see how we can help your business succeed.

💼 Need a raise? | Your dream job is waiting for you! Upgrade your resume, fast-track your job search, and increase your job offers with Teal. Sign up right now – It’s 100% Free! (Sponsored)

📅 LET’S HANG OUT | Upcoming Events
👋🏼 SEE YOU SOON!

🚀 Before you go…Why not join the Explorers Club?

It’s our private Telegram chat where ideas fly, experiments are shared, businesses are born, and we’ve got each other’s backs. Our weekly office hours keep you accountable and spark new ideas. And our network? It is second to none!

Just ask any of our existing members how great The Club is! 

We’re running a promo right now for new members. (It’s normally $35) Trust us… It’s worth it! 😉 

See you inside!

Want to sponsor our newsletter or events? Drop us an email at [email protected]

Join the conversation

or to participate.