🍎Apple’s AI chips are in iPads

ALSO: Shooting victim’s parents are cloning their kid’s voices + celebrity deepfakes

Estimated Read Time: 5 minutes

Apple isn’t ready to reveal all of their AI plans (until June), but they are showing off some of iPad’s upcoming AI features. And unfortunately AI is about to gain some serious attention for ethical concerns as parents clone their dead children’s voices for political messages.

• 💻Apple’s M4 chips are AI-ready.

• 😳Google releases an AI that predicts proteins, can help cure diseases.

• 😱Parents of shooting victims cloned their dead kids to send a message.

• 🎥Can you tell which celebrity is an AI deepfake?

💻Apple’s M4 chips are AI-ready

What happened: Apple isn’t ready to show off their full AI plans yet, but they’ve still revealed their M4 chips, complete with enhanced AI capabilities, coming to the next generation of iPads.

The details: 

• The iPad Pro is getting the M4 chip, with innate support for machine learning features.

• Even the iPad Air, with an M2 chip, was touted as having “machine learning” capabilities such as visual lookup for object identification in photos. (Seems more basic but still an upgrade.)

• Upcoming IOS updates will give developers access to CoreML, allowing you to create core AI functions directly on the iPad.

• Apple is expected to reveal their full AI plans this June at the Worldwide Developer Conference.

Why it matters: The big push here is to get AI locally on your devices. Your computer and phone will eventually handle most simple AI tasks locally, while the cloud will handle more complicated assignments. In other words… AI features we’ve been using through the cloud will eventually be an everyday feature available locally on your device. And it’ll become part of everyday life.

As of market close May 08, 2024.

🔥 Meta’s AI tools allow advertisers to generate full images now. You can even change product colors and angles. The images can be customized with text overlays and adapted to different formats like Reels and Feed on Facebook and Instagram. Pretty cool. But how far can this go before we get into false advertising territory? (Somehow feels like Futurama tried to warn us about this…)

🕵️ Microsoft has developed a “spying” AI for the US intelligence agencies and operates entirely offline. This model is designed to securely process top-secret information without the risk of data leaks. It will be used exclusively by the US government, with no public access. Wonder what the plot of the next James Bond will be now? 🤔

😳Google releases an AI that predicts proteins, can help cure diseases

What happened: Google DeepMind has unveiled AlphaFold 3, the latest iteration of its groundbreaking protein structure prediction model. It’s a massive step forward in AI creating cures to diseases.

The details:

• A huge boost in capability: AlphaFold 3 extends beyond simple protein folding to predict how proteins interact with other important biomolecules, including DNA and RNA.

• This allows AlphaFold to predict complex biology and unlock the secrets to creating new cures and treatments.

• A limited version of the model is online free to use for non-commercial use.

• Proteins are fundamental to all biology, so AlphaFold serves as an important foundation for research.

• Older versions of AlphaFold (since 2018) have been the leading way of predicting protein structure from the sequence of amino acids that make them up.

Why it matters: The ability to predict complex biology is one of the most important features of AI. If AI stays on this course, cures to the world’s worst diseases can become a real possibility.

😱Parents of shooting victims cloned their dead kids to send a message

What happened: The parents of shooting victims called Congress to endorse stricter gun control laws. Except the voices on these calls were… dead kids.

The details: 

• Parents of shooting victims used AI to clone the voices of (now dead) children, who then called and spoke to US Congress.

• It’s part of a new initiative called Shotline.

• The Shotline allows users to send AI-generated voice messages to their Congressional representatives using the voices of children who died to gun violence.

• The whole point is to encourage Congress to enact greater gun restriction laws.

• Shotline uses Elevenlab’s AI voice tech.

• THe voices are generated with the consent of the parents.

Why this matters: This will become a major topic in coming years. Drake already raised Tupac from the dead for his latest song. But you can’t get consent from a dead person. Personally, I don’t care about the consent of the parents when the person in question is gone. It’s a creepy… and emotionally manipulative way of getting the message across. Regardless of the message itself.

This is bringing up massive AI ethics concerns. Trust when I say you’ll hear more about stuff like this soon.

🧐OpenAI shows behind the curtain of their AI

What happened: OpenAI has released “Model Spec”, a framework designed to shape the behavior of their AIs, including ChatGPT. The point of the AI is to create an ethical guideline for the company and use public feedback to alter course.

The details:

• Model Spec is a framework which is designed to define and guide the behavior of their AI models and APIs.

• Much of this appears to be setting an ethical standard for AI. This may affect more than OpenAI if it’s successfully implemented.

• OpenAI is looking for feedback, through a form on their website, in order to improve the framework.

• This comes as part of OpenAI’s recent attempts to be… more “open”. Including their recent

• The Model Spec includes bigger issues such as the potential for AI to generate phishing emails for security training versus its misuse by scammers. (A real concern btw.)

Takeaway: This seems like a good step forward for a standardized ethical framework for AI. The problem is… this is enforced by the company, not the law. Which allows companies to break these rules if they result in more profit. These kinds of ethical frameworks should be enforced onto companies by law, not by choice.

🎥Can you tell which is an AI deepfake?

What happened: The internet went abuzz when these images surfaced… revealing they are indeed deepfakes. For most people, this was the first time they saw how realistic and scary AI generated images can be. For some of you, this will come as no surprise. But most people haven’t seen what AI can do. This was a wakeup call.

These images are theoretical imaginings of celebrities at the Met Gala. And the internet thought they were real…

Takeaway: AI deepfakes are getting crazy. Soon every celebrity and politician will have deepfakes of them around. No deepfake-detector will stop this. And these images served as a wakeup call for the average social media user. This is one of the biggest dangers of AI deepfakes.

🌙 Deepfakes vs Reality - Which ex-president is real?

Hint: None of them were real. Yay…

Source: Midjourney

Give us your feedback!

Got feedback? Rate today’s newsletter by clicking below! 

Got more personal feedback? Fill in the box when you rate us and give us your criticism and feedback.

Thank you for reading! 

❤️Share the (ARTIFICIAL) Love!

Got a friend who wants to appear as smart as you? An AI fanatic or someone shaking in their boots of getting replaced? Share us by clicking the link below and you’ll get our love ;)