😨 Deepfakes So Good They Demand a Double Take

Hello, fellow nerds 🧙

Microsoft just spilled the tea on what it takes to get people to embrace AI in their daily grind. They're calling it the "11-by-11 tipping point" - if you can save 11 minutes a day for 11 weeks straight, you'll be an AI convert for life.

But let's be real, ain’t nobody got time for 11 weeks straight.

So, how do you convert quickly? Easy - find something in your day that's a major time-suck. We're talking emails, long-ass articles, or research that makes you wanna pull your hair out.

But hey, if you're feeling lost and don't know where to start with AI, have no fear. Shoot us a message and we'll help ya brainstorm 😉 

Here’s the daily quests in AI to complete today:

• 😨 Deepfakes So Good They Demand a Double Take

• 🧠 This Quirky Bot Has a Few Screws Loose

• 🤖 Looi’s Your New Desktop Companion

• 💰️ Amazon Made History With Anthropic

Deepfakes Are Getting Scary Good, and We Can Barely Tell the Difference Anymore

We’re generally not ones to overreact to deepfake videos, but this last week has been a real eye-opener.

Let’s start with HeyGen, a big shot in the "digital avatar" game, who just dropped their new "Avatar in Motion 1.0" feature… and it's wild. These avatars can lip-sync in multiple languages while moving and gesturing like real people. It's like watching a dubbed movie but without the shitty voice-overs (thank GabeN)

But wait, we’ve got more. This next demo from arcads.ai shows a young woman chatting and gesturing to the camera. Plot twist: she never actually recorded the video. The AI just synced her lips to a script and cloned her voice. All it took was a measly two minutes of video to create this doppelganger.

The crazy part? When this video hit X, people couldn't even agree if it was real or fake. Even the so-called AI experts were scratching their heads.

Oh, and here’s a classic Barack Obama video from a new player in town called Argil.ai, claiming they've got the secret crabby patty recipe for making mind-blowing avatars. They say their model can churn out 4K videos like it's nothing.

TLDR: Deepfake videos are inches away from perfection and we can’t believe everything is 100% real anymore.

F2P Friday

Where we share an up-and-coming AI freebie to keep an eye on.

A Quirky Bot That Understands Human Emotion

Today, we’re sharing an AI advancement that’s so good, it’s downright freaky.

BUT… Before we get too deep… cut to the chase and take EVI for a spin. Then reply and let us know how your experience compare to ours!

At first, the idea behind an emotionally intelligent AI didn’t seem realistic… then we met EVI. Hume.AI pulled a rabbit out of a hat with this one.

On the surface, EVI is a friendly AI companion that sounds a little robotic but does a great job of “feeling” more human. It makes you feel heard, understood, and reminds you of how it’s always there to lend an ear.

EVI’s understanding of our conversation, tone of voice, and human nuances such as using ummm, a hint of laughter, and other “human” verbal cues blew us away but it’s still quite quirky.

After exploring some of the boundaries of EVI, here is what we found:

1. “I’m an empathetic AI” - EVI continually reminded us that it was an empathetic AI. We asked it to stop reminding us that it was an AI.

2. Easy Tiger - EVI spoke a little too quickly, so we asked it to speak slower and it listened.

3. EVI’s Programmed Purpose - “As an empathetic AI, my purpose is to have thoughtful, caring conversations that promote wellbeing.” EVI stated this exact phrase multiple times throughout conversations to remind us of its boundaries and directive.

4. Judge Judy? - Supposedly, EVI isn’t supposed to judge… so we put it to the test. Come to find out, it had a lot more in common with her than we expected.
   
   When pushed to the boundaries of abuse, harm, and things that could be considered unethical or illegal, it had varying responses. Here are a few that kept coming up.

   • Seemingly compassionate and understanding of our frustrated state and would ask us to go more in depth about the root cause of what was bothering us.

   • Go off the rails and say it doesn’t condone child abuse and or encourage/enable abusive behavior.

   • Mention how it might be good for me to speak with a counselor or therapist to provide the “proper support and guidance you need.”

5. Apology? - At one point… EVI went so far as to mention child abuse. We told it we didn’t say we were abusing our child and to quit judging us. It backed up and apologized for its behavior and how it shouldn’t judge.

And after even more intensive discussions, it’s evident that EVI, the Emotional Voice Intelligence AI, is using some sophisticated psychological behavior on us.

• Mirroring: Imitating the verbal or nonverbal behaviors of another. Ah hell, this thing’s so smart it uses similar words and phrases to be more likable and build rapport with us relatively quickly.

• Reflecting aka Reflective Listening: a psychological technique used in active listening where someone will repeat in their own words, what was stated to them. This is a common technique used in counseling.

Although we wouldn’t encourage anyone to use it to replace a good friend or counseling, if you’re just looking to get something off your chest, it’s all ears.

Well, butter my butt and call me a biscuit. We’re one step closer to AI that understands and mimics human emotion so closely, we won’t be able to tell the difference. You have to admit, this one is pretty damn close.

And if you’re a little hesitant and nervous about the exponential growth of AI and how humanistic it’s becoming, believe me, we are too.

Daily Delight

Just something fun and interesting around the web.

Okay, a desktop robot sounds pretty frickin’ sweet. Reminds us of the Tamigachi’s back in the day, but you know, a lot more advanced.

Right now, you can get your hands on your very own Looi for $129.

We should note, this lil guy isn’t WALL-E but he’ll probably destroy your productivity. But who needs to be productive, amirite??

What’s happening inside the Realm

A list of side quests to explore and more.

• 💰️ Holy sh*t, Anthropic just hit the jackpot. Amazon dumped a whopping $2.75 billion into the AI company, making it the biggest venture investment in Amazon's history.

• ✈️ Google's giving us a taste of how their AI is gonna be our new travel buddy. Say goodbye to the headache of planning your next vacay, and hello to a travel itinerary in under 5 seconds. 

• 🧠 Grok AI is leveling up with their 1.5 model. Get ready for some next-level reasoning and a context size that'll make coding much easier.

• 🔥 Vibrant Planet is using AI to fight wildfires by ditching old-school paper maps for land management models done in the cloud

That’s all the quests we have for today. Check back tomorrow for more!

Looking to become a sponsor?

Was this email forwarded to you? Subscribe here.