Microsoft’s back-to-school AI

PLUS: AI restores woman's speech

TOGETHER WITH

Good morning, human brains. Welcome back to your daily munch of AI news.

Here’s what’s on the menu today:

  • Meta Introduces SeamlessM4T

    An open multilingual AI translation model.

  • Microsoft’s back-to-school AI 👩‍🏫 📚

    Microsoft gets a new suite of AI features.

  • AI performs a miracle 👩‍🦽 🎉

    A paralyzed woman speaks for the first time in 18 years.

APPETIZER

Meta’s new multilingual AI 🎤 🫦

On Tuesday, Meta unveiled SeamlessM4T. It’s a multilingual AI translation model.

It supports nearly 100 languages.

The juice:

1/ SeamlessM4T is the first open “all-in-one” multilingual, multimodal AI translation, and transcription model.

2/ It supports speech-to-text, speech-to-speech, text-to-speech, and text-to-text translations.

3/ The model’s speech-to-speech translation supports 100 input and 36 output languages.

4/ Users can switch languages mid-sentence in text-to-speech.

Our take: Babel Fish from “The Hitchhiker’s Guide to the Galaxy“ vibes 😎

Meta claims this is an “open” model. They’ve continued to face controversy over their use of the word “open-source” to describe less than open-source releases. Could be an intentional change in wording…

Sneaky, sneaky.

BUZZWORD OF THE DAY

Open-source

The term open source refers to something people can modify and share because its design is publicly accessible.

Open source projects, products, or initiatives embrace and celebrate principles of open exchange, collaborative participation, rapid prototyping, transparency, meritocracy, and community-oriented development.The process of cleaning, transforming, and analyzing large sets of data to discover meaningful patterns, trends, and insights using artificial intelligence and machine learning algorithms.

FROM OUR PARTNERS

Looking for world-class AI & data science devs?

AE Studio is a development, data science, and design studio. They work with founders and executives to create custom software, machine learning, and BCI solutions.

Your brain while working with AE (allegedly).

Whether it’s spinning up an MVP, Enterprise Digital Transformation, or applying AI & ML to your business — AE’s blend of expertise and bleeding-edge pedigree means you’ll be working with the finest.

To be more specific, pedigree is like cutting your teeth in the AI crucibles of Stanford, MIT, Harvard, and Caltech.

And expertise, like producing ROI for Berkshire Hathaway, Point, EVgo, Protocol Labs, and Biocentury.

Lucky for you — AE Studio’s taking on new clients, for a limited time.

MAIN COURSE

Microsoft’s back-to-school AI 👩‍🏫 📚

Microsoft Designer is here for US Edge users. It’s an app that integrates AI-powered graphic design capabilities into the browser.

Edge itself also received several more AI updates.

Microsoft Edge has added smart browser features to enhance creativity, productivity, organization, and learning for the “new school year.”

The main points:

1/ Microsoft Designer — a design tool in Edge’s sidebar, that lets users create and share designs directly on email, social media, etc.

2/ Smart Find — an AI-powered feature that finds content within a web page, suggests synonyms, corrects grammar, etc.

3/ Bing Chat — in its sidebar home, Bing Chat summarizes articles, gives shopping advice, organizes browser tabs, etc.

4/ Microsoft 365 integration — Outlook, Word, OneNote, etc., are now directly accessible through the sidebar.

5/ Tabs — vertical tabs and tab groups allow users to organize tabs more efficiently.

6/ Immersive Reader — “distraction-free“ reading environment with AI grammar tools.

7/ Read Aloud — allows users to listen to written content.

Our take: as of May 2023, 64% of internet users go with Google Chrome, 21% use Apple Safari and 5% use Microsoft Edge. Will these updates increase this number?

We’ve ragged on Edge before, but frankly… all this functionality sounds pretty appealing.

A LITTLE SOMETHING EXTRA

AI restores woman’s speech 👩‍🦽 🎉

A paralyzed woman just spoke for the first time in 18 years with the aid of AI. UCSF & UC Berkeley researchers developed an AI-powered brain implant that has allowed her to speak via a digital avatar.

Behind the miracle:

1/ The implant translates the patient’s brain signals into speech.

2/ An AI animates a digital avatar based on the patient’s wedding video and the generated speech. The avatar is able to mimic her intended facial expressions and tone of voice.

3/ It transcribes her speech into text at around 80 words per minute.

Our take: this advancement in BCI (brain-machine interface) tech hints at a future of natural communication for the paralyzed. We’re certain there are more breakthroughs to come before the end of the year — Neuralink was approved for human testing this last June.

MEMES FOR DESSERT

YOUR DAILY MUNCH

Think Pieces

Do robots fly planes better than humans? ”Pibot,” developed by the Korea Advanced Institute of Science & Technology, is a humanoid that flies planes without major cockpit adjustments.

Google and YouTube play both sides. YouTube partners with Universal Music Group, while Google scrapes the entire web to train its AI models for free.

India uses AI-powered voice recognition payments. The Reserve Bank of India will enable users to verbally give transfer instructions via their phones.

Startup News

Meta announced an EU “AI off switch” for social media users — an option to get rid of the AI & personalization timelines, replacing it with a chronological or posts based on local popularity.

Hugging Face has made an open-source replication of Flamingo. Hugging Face used Llama v1 and CLIP to create a “comparable” replica of DeepMind’s state-of-the-art vision-language model.

NVIDIA and VMware expand partnership. The goal: assist companies in training and integrating generative AI models.

Research

Google Research paper on enhanced robotic learning via rewards. A method to teach robots new actions via natural language input.

Giraffe — a method to extend context lengths of LLMs past their training limits. Also compares RoPE, ALiBi, xPos, linear scaling, and other methods for zero-shot context extrapolation.

Cybersecurity LLMs — a paper comparing how ChatGPT, Bard, and Microsoft Bing solve CTF (Capture-The-Flag) challenges and answer professional certification questions.

Tools

Recast [Sponsored] — turn any article into a concise audio podcast. Your AI co-hosts save you time by explaining all the important details with engaging dialogue.

AI Office — allows users to create a unique version of AI Town.

Gamma 1.0 — a visual presentation creation tool with AI-powered text, visuals, images, layouts, etc.

TWEET OF THE DAY

A new generative AI startup from Google Brain alums.

Tag us on Twitter @BotEatBrain for a chance to be featured here tomorrow.

RECOMMENDED READING

🏥 Healthcare AI News — 5 minutes or less is all it takes to elevate your Healthcare knowledge with this expert-curated weekly AI newsletter. Stay informed and stay ahead.

👨 The Average Joe — Market insights, trends and analysis to help you become a better investor. We like their easy-to-read articles that cut right to the meaty bits.

💌 Marketing Letter — The newsletter keeping 30k+ marketers in the loop. Read by marketers who work at LinkedIn, Techcrunch, and Disney.

AI ART-SHOW

Until next time 🤖😋🧠

What'd you think of today's newsletter?

Login or Subscribe to participate in polls.