🤖 This week’s AI news

ALSO: Google reveals their voiced AI

Estimated Read Time: 5 minutes

The AI race heats up as Google reveals Gemini Live, their equivalent of GPT-4o’s voice mode. And Grok 2 is officially a state-of-the-art AI model. Robot arms, Walmart bragging about productivity, and coders replacing themselves. Let’s dive in!

• 📢 Google beat OpenAI in the “Voice” race.

• 📈 Grok 2 is out… and it’s impressive.

• 🦾 iPad is getting a robot arm.

• 🤖 Autonomous coders and scientists are incoming.

• 💰 Walmart brags they use AI to increase productivity.

📢 Google beat OpenAI in the “Voice” race

What happened: Google just released Gemini Live, their voiced-AI with natural conversation ability, for limited testing. It competes directly with GPT-4o’s voice mode, which hasn’t released yet. It’s also the built-in assistant for the Pixel 9 phone.

The details: 

• Gemini Live features advanced hands-free voice conversations using 10 different realistic voices.

• You can interrupt Gemini Live and it’ll respond accordingly like in a real conversation.

• However Gemini Live cannot see or understand your camera like GPT-4o’s promised advanced features.

• Gemini will integrate directly with Google to understand the context of your requests… just like GPT-4o’s advanced features.

Why it matters: Gemini Live is an impressive step forward for voiced AI. I suggest you try it out. It’s quite impressive. And it goes to show that Google is staying competitive with OpenAI. Once these models get into our phones and computers in a natural way, they’ll change how we interact with our devices.

🚗 New AI robots are already building cars in factories. Figure AI just showed off their new robot working a car manufacturing environment. It’s impressive stuff. Check out the video here.

📸 OpenAI is investing $60M in webcams company Opal. Opal’s plan is to move beyond mere webcams and into AI-integrated tools using OpenAI’s tech. Reportedly OpenAI is interested in using their voice tech with Opal’s tech.

📝 Google Meet’s getting a new AI note taking feature. Now Gemini will automatically take notes from your online Google Meeting, reducing the need for manual note taking. It’ll be available to Workspace customers, with the AI feature costing $10 a month for Workspace members.

📞 The FCC cracked down on AI voice calls by suggesting a new law which would require AI-powered voice calls to disclose they are AI-powered. It’s easy to spam a million people with robocalls. It’s also easy to scam people by faking the voice of somebody they know or a celebrity. This law was guaranteed to happen. The only problem is enforcing it.

📈 Grok 2’s performance is impressive

What happened: Grok 2 is now available for beta users on X/Twitter. And it’s boasting some impressive performance; outperforming both Claude and ChatGPT.

The details:

• Grok 2 mini is also available on X and will have an enterprise version available later.

• Grok 2 can now create and publish images directly on X. Testing will see how good the images are.

• Performance metrics show Grok 2 surpasses Claude 3.5 and GPT-4o in multiple areas, making this a cutting edge AI model.

Why it matters: The more competition, the better. But Grok 1 was also open sourced, and I’m hoping Grok 2 will be as well. The only issue was that Grok 1 was released at an unusable weight, making it a rather insincere open source release. Grok 2 becoming open source in a useable form, combined with Llama 3’s release, would be a massive step up for free, open source AI.

🦾 iPad is getting a robot arm

What happened: Apple is reportedly developing a tabletop smart home device with robotic arm, a built in display and Siri voice features. We’re in the future now.

The details: 

• The project is codenamed J595 (wow) and is aimed to launch by 2027, possibly earlier.

• It’ll be a tabletop smart-home device with a robot arm, iPad-like display, and Siri voice features.

• The robot arm will reportedly grab the thin iPad-like display and move it around with full rotation.

• It’s expected to cost at least $1,000 and run on iPadOS.

Why this matters: Apple’s foray into any industry is a landmark moment because they affect public perception. The Vision Pro has opened up the public to VR headsets

🤖 Autonomous coders and scientists are incoming

What happened: Fully autonomous AI scientists and coders are real. Companies are trying to automate even the most difficult tasks.

The details:

• Japan-based Sakana AI introduced their AI scientist, supposedly capable of fully automating scientific research. Needless to say this could revolutionize science and create new breakthroughs.

• Creating a research paper with Sakana costs around $15 to create, bringing scientific research to a wider audience.

• Sakana goes one step further. They believe someday AI will have its own review boards and hold its own conferences.

ALSO:

• Cosine revealed Genie, their autonomous software engineer. It’s powerful enough to break the performance record for coding for LLMs.

• Genie is trained on a dataset made from actual work by coders, so it should theoretically program more like a human would.

• A big differentiator: Genie will recalculate its actions after a mistake and attempt to fix the issue. This is something historically difficult for most other AI models to achieve.

Takeaway: Coders are pretty eager to replace themselves. But an autonomous scientist is big news and could potentially revolutionize scientific breakthroughs. But every company is eager to hype themselves up. Let’s see how these AI models actually compare in real life.

💰 Walmart brags they use AI to increase productivity

What happened: Walmart’s CEO McMillon just bragged how much AI has boosted his company’s productivity during a Q2 earnings call. Take a glimpse at the future.

The details:

• Walmart is using AI to update millions of product entries. Apparently it would require 100 times the workforce to do the same job.

• This includes using AI to generate product images to help employees quickly identify items for online orders.

• Walmart is also introducing a new AI shopping assistant on its website and app to offer personalized shopping advice to customers.

Takeaway: They're already getting this much of a benefit from basic AI processes. I imagine AGI will revolutionize the workforce (and cost people a lot of jobs). This is the direction things are going in.

🌙 Supernatural portraits…

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 ;)