- Everyday AI
- Posts
- ChatGPT’s New Advanced Voice Mode: 5 things you need to know and live demos
ChatGPT’s New Advanced Voice Mode: 5 things you need to know and live demos
Meta Llama 3.2, ChatGPT Advanced Voice mode explained, EU AI pact, Microsoft's new correction feature and more!
Everyday AI
September 25, 2024
👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI
Outsmart The Future
Sup y’all 👋
Been a wild few days of updates, right?
In the past few days, we’ve seen MAJOR updates from OpenAI, Google AND Meta.
Yikes, what a week.
Today’s episode was actually a ton of work. (I’m kinda sleepy TBH.)
But our team really put in the work on today’s episode, as what OpenAI released could really change how…. Society works?
Hope you can give today’s show a watch/listen/read.
✌️
Jordan
Today in Everyday AI
7 minute read
🎙 Daily Podcast Episode: ChatGPT just released its new Advanced Voice mode. We break down what you need to know and give a live demo. Give it a listen.
🕵️♂️ Fresh Finds: A new multimodal model enters the ring, Reddit expands AI-powered summaries and Max adds Google AI closed captioning. Read on for Fresh Finds.
🗞 Byte Sized Daily AI News: Meta unveils Llama 3.2, EU unveils AI pact members and Microsoft launches new correction. For that and more, read on for Byte Sized News.
🚀 AI In 5: ChatGPT’s new Advanced Voice mode is now live! We noticed there’s one major flaw that you should be aware of. See it here
🧠 Learn & Leveraging AI: Here are the 5 things you need to know about ChatGPT’s new Advanced Voice Model. Keep reading for that!
↩️ Don’t miss out: Did you miss our last newsletter? We talked about OpenAI unveiling Advanced Voice mode, Gemini added to Workspace and Russia using AI to interfere with the US election. Check it here!
ChatGPT’s New Advanced Voice Mode - 5 things you need to know and live demos 🗣
So good it's literally illegal?
ChatGPT's new Advanced Voice Mode just dropped to all paid users. (Well, except for countries where it's kinda technically not legal.)
We're gonna do live demos, take some LIVE user requests, and tell you the 5 things you need to know about this new groundbreaking feature.
Join the conversation and ask Jordan questions on ChatGPT here.
Also on the pod today:
• Advanced Voice Mode Capabilities 💬
• Current Limitations of Advanced Voice Mode 🤔
• Live Demo of Advanced Voice Mode 🧑💻️
It’ll be worth your 55 minutes:
Listen on our site:
Click to listen
Subscribe and listen on your favorite podcast platform
Listen on:
Here’s our favorite AI finds from across the web:
New AI Tool Spotlight – Polymet is an AI product designer, Magic Inspector is an AI web test automation platform and Nuvio provides AI-powered financial management.
Trending in AI – AI2’s new multimodal model Molmo is giving other big name multimodal models some major competition.
Social Media – Reddit bringing AI-powered automatic translations to a bunch of new countries.
AI in Media - Max is getting Google AI-generated close captions.
AI Design - Figma’s AI-powered app generator is once again available after its Apple copyright situation.
Money in AI – Nebius Group, emerging from a deal to acquire Yandex's Russia-based assets, plans to invest over $1 billion in AI infrastructure across Europe.
Read This – Some video game studios are turning to AI to make NPCs more interactive.
1. Meta Launches Llama 3.2 with Enhanced AI Models 👀
Meta has unveiled Llama 3.2, featuring new lightweight models and advanced vision capabilities designed for broader accessibility. The release includes small and medium-sized models that can operate on edge devices, making it easier for developers without extensive resources to harness powerful AI tools.
With the ability to reason with images and generate multilingual text, Llama 3.2 is set to empower app development while ensuring user privacy through local processing.
2. EU Unveils AI Pact Signatories: Apple and Meta Missing 👀
The European Commission has unveiled its first batch of over 100 signatories to the AI Pact, aimed at encouraging companies to make voluntary commitments towards responsible AI usage while navigating the upcoming AI Act's compliance deadlines.
With major players like Amazon and Microsoft on board, the initiative seeks to foster collaboration and information sharing among signatories, though notable absences like Apple and Meta raise eyebrows regarding their approach to compliance. As companies gear up for potential penalties that could reach billions, this pact might just become the new benchmark for AI accountability in Europe.
3. Microsoft Launches New “Correction” Feature in Azure AI 🛠️
Microsoft has introduced a new “correction” feature in Azure AI Studio, designed to automatically identify and amend inaccuracies in real time. This tool scans AI-generated content against source materials, flagging errors and providing corrections before users even see the mistakes.
However, while this initiative aims to tackle the common issue of AI "hallucinations," it doesn't guarantee complete accuracy and may still produce errors, as noted by Microsoft representatives.
4. Google Rehires AI Pioneer Noam Shazeer 👤
Google has resecured the talents of Noam Shazeer, a key figure in the AI boom known for co-authoring a groundbreaking research paper. After leaving Google in 2021 to launch Character.AI, Shazeer's startup struggled to gain traction, prompting his former employer to swoop in and bring him back.
This development not only highlights Google's commitment to leading the AI landscape but also underscores the challenges startups face in this competitive arena.
5. AI Uncovers Hidden Treasures in Nazca Desert 🏺
In a groundbreaking study, researchers from Yamagata University and IBM have harnessed artificial intelligence to nearly double the number of known geoglyphs in Peru's Nazca Desert, revealing 303 new figures that include humans, animals, and even a knife-wielding killer whale.
This innovative AI model analyzed drone-captured images to detect faint outlines that would otherwise remain hidden, showcasing the technology's potential in archaeological discoveries.
6. Microsoft Bets Big on Mexico's AI Future 🇲🇽
Microsoft AI Tour 2024 in Mexico City, CEO Satya Nadella revealed a staggering $1.3 billion investment over the next three years to boost AI infrastructure and skills in Mexico. This initiative aims to democratize AI access for 5 million people, particularly benefitting small and medium-sized businesses looking to enhance their digital capabilities.
With companies like Grupo Bimbo and Cemex already leveraging AI for competitive advantage, this investment could significantly reshape the tech landscape, making it easier for individuals and organizations to harness the power of AI for career growth and operational efficiency.
One Fatal Flaw of Advanced Voice Mode Inside ChatGPT
ChatGPT has just released its highly anticipated Advanced Voice Mode and it’s impressive!
BUT there’s one major flaw that no one is talking about.
We break down what you need to know about this new feature.
Check out today's AI in 5.
🦾How You Can Leverage:
Welp, this new voice thing is kinda weird.
In a good (and maybe bad?) way.
About 24 hours ago, OpenAI started releasing its Advanced Voice Mode to most paid users.
And things in the AI space will probably never be the same.
ChatGPT's new advanced voice mode (AVM) isn't just another Alexa clone – it's the cool, emotionally intelligent cousin that makes Siri look like it's still using dial-up.
It TOOOOOOOTALLY is, right? Lolz.
We took this new verbal virtuoso for a live spin on today’s show, and holy chatbot, Batman!
It's like having a stand-up comedian, a polyglot, and a Fortune 500 consultant BFF rolled into one AI package.
But before you start planning your AI karaoke night, let's break down the five things you absolutely need to know:
Let’s goooooooo
1 – Exclusive Club: Paid Members Only 💸
ChatGPT's fancy new voice is playing hard to get. It's only available to paid ChatGPT Plus and Team users. Free users? Prolly never. Enterprise? You'll have to wait your turn.
Within the next week, most Plus/Teams accounts should have access. (Except for people in #5 below. Yikes)
What it means:
OpenAI is clearly targeting the cream of the crop, probably to manage server load and gather primo feedback.
It's like the AI equivalent of a velvet rope and a bouncer. If you're not on the list, you're not getting in... yet.
2 – More Guardrails Than a Toddler's Playground 🛝
No singing, no stuttering, no celebs impersonations, and definitely no falling in love.
It's like OpenAI gave it strict AI etiquette lessons.
What it means:
OpenAI is playing it safe, avoiding potential PR nightmares and legal headaches.
We get it.
Jailbreaking a model via text input/output is one thing. Getting it to SPEAK outside of the guardrails is another.
But don't worry, there's still plenty of room for fun.
You just might have to get creative with your requests. Who needs AI love when you can have AI sass? Or AI minions teaching you in Spanish. It’s fun.
3 – The Usain Bolt of Voice Chats 🏃
So fast.
Like…… talking to a human fast.
Again, this isn't your grandma's voice assistant that literally knows nothing and has a 3-5 second delay between each response.
It's faster, smoother, and more intuitive than anything else on the market. Plus, it's available on both iPhone and Android, making it the widest available AI smart assistant right now.
Oh, and this is important to note.
The standard voice mode in ChatGPT has been available for months. Compared to the new AVM, the standard voice mode is kinda like talking to a pile of bricks.
(At least in terms of experience.)
What it means:
We're entering a new era of human-AI interaction.
Seriously. Might seem weird now, but remember how weird it was to sit in front of a computer all day in the 90s? Yeah, same feeling.
Real-time voice assistants with emotions could change everything from customer service to personal productivity.
Imagine having a lightning-fast, emotionally intelligent assistant always ready to chat. It's like having a super-smart friend who never sleeps.
The real benefit here? One of the biggest challenges with ai right now is…. Showing ROI.
How about this?
Talking vs. Typing
Average Talking Speed: Approximately 150 words per minute (wpm)
Average Typing Speed: Approximately 40 words per minute (wpm)
All of a sudden, your ceiling for productivity just went up almost 4X.
Sheeeeeeeeeeesh.
4 – Know it all. Not a do it all. 🤔
This part’s a HUGE bummer.
(But we’ll have a video tomorrow to address this.)
Here's the kicker: when you're in advanced voice mode, all other ChatGPT tools go bye-bye. No internet browsing, no GPTs, no typing. It's voice or nothing, baby.
If you start a new chat in AVM and then do anything else, that chat then loses the AVM ability. (It’ll default back to standard voice mode.)
So kinda like the o1 model, Advanced Voice Mode right now works in a silo.
What it means:
This is clearly version 1.0.
OpenAI is focusing on nailing the voice interaction before integrating other features. It's a bit like having a Ferrari that can only drive on one street. Awesome, but limited. Expect this to change as the tech evolves.
5 – Giving EU Regulators Night Sweats 😰
If you're in the UK, EU, or a few other countries, you're out of luck.
Sorry.
This voice mode might actually be temporarily illegal in EU schools and workplaces. Apparently, emotion detection is a big no-no.
Although, OpenAI and the EU did just kinda shake hands today, so we’re guessing this will change sooner rather than later.
What it means:
We're witnessing the clash between rapid AI advancement and cautious regulation in real-time. OpenAI might need to create a special "EU version" that's more vanilla.
It's like making a decaf version of Red Bull – it kind of defeats the purpose.
⌚
Numbers to watch
105 minutes
Customers using Gemini for Google Workspace save an average of 105 minutes per user, per week.
Now This …
Let us know your thoughts!
Vote to see live results
In our AI in 5 segment on YouTube, we often create different types of videos. Which are your favorites? |
Reply