HubSpot Inc.

10/01/2024 | News release | Distributed by Public on 10/01/2024 02:40

Monthly AI Roundup: The Biggest AI Headlines in September

Monthly AI Roundup: The Biggest AI Headlines in September

Published: October 01, 2024

There's so much AI news. Our monthly roundup pulls together some of the biggest headlines so you don't have to.

This month: Notebook LM's audio feature, all things Apple Intelligence, AI sunglasses, and more.

Tools & Features You Should Know About

Editor's Pick: Notebook LM's new audio feature is beyond impressive.

If there's one update you should know about, it's this.

Google's Notebook LM has been around for a year and last week, its newest audio feature had me sitting in disbelief with my mouth agape. In all my time covering AI, that has never happened.

Notebook LM's audio feature, released on Sept. 11, allows you to listen to two hosts have an incredibly natural conversation about the content you upload to the platform.

All you have to do is upload your files, wait a few minutes, and BAM: You'll have two AI hosts discussing your document in incredible detail, with banter and jokes. I'd go as far as saying it's nearly undetectable from real people, no uncanny valley.

Don't believe me? I uploaded this localization article to Notebook LM. Here's the 10-minute breakdown it provided.

Play audioPause audio

00:00/00:00

Show transcript Hide transcript

The output from GenAI, thus far, has always been a bit off. And this is across mediums: Video, text, and audio.

Even audio I've generated on ElevenLabs haven't sounded this good. They've lacked the proper cadence, pauses, and natural flow of real conversation.

Powered by Gemini 1.5, Notebook LM nails it.

Potential Use Case: If you're an auditory learner, this tool is for you. Students on TikTok say this is transforming how they study and learn complex information.

Here are its limitations:

  • The hosts can only process and deliver English content.
  • You currently do not have the ability to edit the audio content.
  • There's always a risk of hallucinations, although the risk should theoretically be lower.

Meta AI gets exciting updates.

Mark Zuckerberg and company have been busy.

Meta's Ray-Ban smart glasses just got some pretty cool upgrades. You can now ask Meta AI to:

  • Send reminders about everything from where you parked to sending that "I'm home" text after a night out.
  • Give you information about landmarks, restaurants, and stores as you walk through a city.
  • Give live language transition to/from French, Italian and Spanish.
  • Connect to music streaming apps like Audible and Amazon Music.

Also, if you're using Meta's AI assistant on WhatsApp, Instagram, and Facebook, you'll notice the addition of celebrity voices, like Kristen Bell, John Cena, and Judi Dench, who will voice all interactions you have.

Video Source

Apple Intelligence

In October, Apple's expected to release its first batch of Apple Intelligence features with iOS 18.

Image Source

Here's what's coming:

  • Siri's getting a huge upgrade - The assistant will be able to answer more complex questions, including follow-ups, and will integrate with ChatGPT.
  • Custom emojis - Want an emoji of a woman holding a black cat while eating pizza? Oh, just me then. Well, I'll be able to do that soon.
  • Call, text, email, and notification summaries - Want to get quirk blurbs about your most important conversations? Apple Intelligence will have you covered.
  • Visual intelligence - Your camera can now offer real-time support based on what you show it, like identifying landmarks, listing store hours, and more.

For some of these features, consumers will have to wait until early next year. Hopefully, good things actually do come to those who wait.

Here's What Big Tech's Up To

Open AI and Its Lone Wolf Sam

OpenAI is going through quite the transition phase.

On Sept. 25, CTO Mira Murati shared on X that after six years, she was leaving the company for her "own exploration." Shortly after, VP of Research Barret Zoph and Chief Research Officer Bob McGrew announced their departures.

Here's a breakdown of the departure streak that started late last year:

  • November '23 - Former board members Helen Toner and Tasha McCauley leave after a failed attempt to oust COE Sam Altman.
  • February '24 - Research scientist Andrej Karpathy leaves and later founded an AI education startup.
  • May '24 - Chief Scientist and co-founder Ilya Sutskever, who was rumored to lead the ouster against Altman, leaves and starts his own AI company Safe Superintelligence.
  • August '24 - Co-founder John Schulman leaves and joins Anthropic, the creator of the AI chatbot Claude.
  • August '24 - Though not technically a departure, President Greg Brockman announces a sabbatical through the end of the year.

Image Source

In other - potentially related - news, OpenAI is rumored to transition from a nonprofit org to for-profit. Critics say this is just them moving more toward the direction they've been headed in for a while now.

ICYMI, OpenAI has long been criticized for prioritizing profit over safety in its AI pursuits. This news, if true, will heighten those fears and puts into question the company's mission of developing safe AI.

The recent batch of exits may signal that former employees feel the same way.

The FTC takes on 'deceptive AI claims.'

We've all heard some of the nefarious ways AI can be used. Well, the Federal Trade Commission (FTC) is cracking down on companies leveraging AI to encourage deceptive tactics.

Last week, the FTC announced an investigation into multiple companies as part of their "Operation AI Comply" sweep.

Here are some companies under fire:

  • DoNotPay, a company claiming to have the world's first robot lawyer, for giving wildly inaccurate information.
  • Ascend Ecom for falsely promising AI-powered tools to help consumers start their ecommerce platforms. $25 million later, consumers never got what they were promised.
  • Rye for having an AI writing assistant that would generate fake reviews and testimonials.

We've covered the allure of having the buzzword "AI" in your marketing materials. The moral of the story is: Make sure your product lives up to the hype. Oh, and don't leverage AI to do something illegal.

There you have it - the top news you should know about. We'll catch you next month!

Don't forget to share this post!