• NATURAL 20
  • Posts
  • Ive & Altman Team Up for AI Devices

Ive & Altman Team Up for AI Devices

Apple's Design Chief Collaborates with OpenAI's Sam Altman to Forge New Frontiers in AI Hardware

Today:

Apple’s iPhone Design Chief Enlisted by Jony Ive, Sam Altman to Work on AI Devices

Tang Tan, a key figure in Apple's iPhone and Apple Watch design, is set to leave the company in February to embark on a new venture in artificial intelligence hardware with LoveFrom, a design firm led by Jony Ive. Ive, renowned for his role in designing some of Apple's most iconic products, left Apple in 2019 to establish LoveFrom. This new project, still in its early stages, aims to develop AI hardware, with OpenAI's CEO Sam Altman providing software support.

This collaboration is one of Ive's most ambitious undertakings since his departure from Apple. Tan will lead hardware engineering for the project, marking a significant shift in his career after stepping down as Apple's vice president of iPhone and watch product design. The project is currently focusing on hiring talent and developing initial concepts, including potential devices for home use.

Musk’s xAI Incorporates as Benefit Corporation With ‘Positive Impact’ Goal

Elon Musk's new AI company, xAI, is making waves by choosing a unique corporate structure. Incorporated in Nevada in March 2023, xAI stands out as a for-profit benefit corporation. This status lets the company prioritize social impact over shareholder obligations. It's a move that aligns with rivals OpenAI and Anthropic, showcasing a trend in the AI sector towards socially responsible business models​​​​.

xAI, while keeping details under wraps, has big ambitions to "understand the true nature of the universe." Musk, known for his involvement with SpaceX and Tesla, has gathered a team including veterans from notable AI companies like OpenAI and DeepMind, and intends to collaborate closely with his other ventures, including Tesla and Twitter (now X Corp.). The company is set to explore advanced AI capabilities, with a focus on ethical and safe development. This approach reflects a growing awareness in the tech world about the potential risks and responsibilities associated with AI technology​​.

It's a bold step for Musk, who's been a critical voice in the AI field, especially after his departure from OpenAI's board in 2018. With xAI, he's charting a new path in the AI landscape, balancing innovation with social responsibility.

Seamless Communication | AI research by Meta

Meta's AI research has rolled out a game-changing suite of translation models, known as Seamless Communication. This lineup includes four key components: Seamless, SeamlessExpressive, SeamlessStreaming, and SeamlessM4T v2. They've managed to blend high-quality multilingual translations with quick response times and the unique ability to keep the speaker's original vocal style and expressiveness, even during live translations. 

SeamlessExpressive can transfer not just words, but the speaker's emotional vibes and style across different languages. They've also developed SeamlessAlignExpressive, a massive collection of audio clips in multiple languages, all lined up to match in meaning and mood. 

SeamlessStreaming is another star player, smart enough to figure out the perfect moment to start translating, adapting to different languages' quirks. And SeamlessM4T v2? It's a powerhouse, trained on a whopping 4.5 million hours of speech, making it super accurate and versatile.

Microsoft is training an AI to help get nuclear reactors approved

Microsoft's working on an AI to speed up the approval of new nuclear reactors, aiming to power its data centers. These centers are crucial for running OpenAI's ChatGPT, which has a huge user base and needs tons of computing power. 

Microsoft's already deep into OpenAI, investing big bucks and building a monster supercomputer for it. They're eyeing small modular nuclear reactors (SMRs) for clean, reliable energy. 

Right now, getting SMRs approved is a big headache - super costly and complex. To make things easier, they've partnered with Terra Praxis to train an AI for whipping up the necessary paperwork. This move could slash the human effort by 90%, a real game-changer for their nuclear energy plans.

NGA leans in on AI, machine learning to improve data analysis

The National Geospatial-Intelligence Agency (NGA) is going big on AI and machine learning to handle heaps of data real quick. They're using this tech to make sense of satellite images and figure out new ways to manage the data. The human factor is key, especially when dealing with bias in data, keeping the outputs on track, and staying safe tech-wise.

Natasha Krell, an NGA scientist, likens AI to a word processor. It doesn't do your job but helps you do it better and faster. NGA's using AI to spot the needle in the haystack in imagery analysis. AI helps them spot and identify objects, especially with tons of satellite data coming in. They’re exploring both in-house and cloud solutions for storage and computing.

NGA’s got different ways to look at images, like electro optical, SAR, and thermal. Combining these is the next big thing in AI. They're also using AI for humanitarian work and sharing insights with other agencies.

IBM taps AI to translate COBOL code to Java

IBM's rolling out a new tool, Code Assistant for IBM Z, to make old-school COBOL code talk modern Java. COBOL's been around since '59 and is everywhere, but it's a pain to work with. Swapping it out for something newer is usually a headache that costs big bucks. IBM's gizmo, using AI, aims to make this switcheroo smoother and cheaper. It can handle COBOL and Java, plus 80 other languages. 

The tech's smart, using a bunch of data to keep the code's original vibe while changing its language. IBM's betting big on this, especially since most of their mainframe users are still using COBOL. They're not just stopping at Java; they're eyeing bigger AI coding tools, competing with the likes of GitHub Copilot. IBM's all about connecting old tech with new, making it easier for businesses to stay up to date.

For some blind and low-vision people, AI glasses unlock a new independence

Wes Ramage, born with optic nerve hypoplasia, learned to map out new environments despite his limited vision. Now 43, he wears AI-powered Google Glass by Envision, assisting his mobility. These glasses, connected to a large language model like ChatGPT, snap a photo and describe the scene when commanded. This tech has been a game-changer for blind and low-vision individuals, offering detailed information beyond basic object recognition.

While not flawless (sometimes creating or misinterpreting scenes), these advances offer a new level of independence. Ramage, for example, uses his glasses for everyday tasks like reading signs or finding items. Envision's technology, partly based on OpenAI's models, has evolved to narrate surroundings. However, it's not cheap, with prices ranging from $1,899 to $3,499.

The Biggest Of The Big: AI Startups Raised Huge — These Were The Largest Deals Of 2023

AI startups were on fire in 2023, raking in serious cash. Microsoft dropped a whopping $10 billion on OpenAI, signaling their big move in the AI game. Anthropic wasn't far behind, bagging nearly $7 billion, including a $4 billion investment from Amazon. These guys are ChatGPT's main rival and they're using Amazon's tech to up their game.

Inflection AI, another player, scored $1.3 billion with big names like Microsoft and Nvidia backing them. They're cooking up something big with their AI assistant, Pi. Then there's Metropolis, a parking startup that's more than meets the eye, thanks to its AI and computer vision tech. They pulled in $1.1 billion and even made a mega-deal to take SP Plus private.

Databricks, not strictly an AI firm, still made the cut with a cool $685 million, boosting their value to $43 billion. Meanwhile, Aleph Alpha and SandboxAQ each grabbed $500 million, showing that AI's not just a U.S. game. Mistral AI in Paris also made waves with a hefty $487 million round.

Big Tech firms like Microsoft, Nvidia, Amazon, and Google were the ones opening their wallets. Nvidia, especially, is killing it with its AI investments. It's clear that these tech giants are driving the AI investment scene, leaving traditional venture firms playing catch-up. 2023 was a banner year for AI funding, and it looks like this trend isn't slowing down anytime soon.

Microsoft CEO Satya Nadella discusses the promise and potential perils of AI

Microsoft CEO Satya Nadella chats about the real deal with AI in a recent NPR interview. He reckons 2023's the year AI's really kicking off, becoming a big part of our everyday life. Microsoft's teamed up with OpenAI, the folks behind ChatGPT, and they're slipping AI features into stuff we use all the time, like Excel and Word. Nadella's all about this "copilot" idea, where AI helps out but doesn't take over. He's clear that AI's not perfect and we gotta stay sharp when using it.

Nadella's also got his eye on the bigger picture, making sure AI's used for good and not for sketchy stuff like bomb-making. He's aware that Microsoft's in a powerful spot, controlling what info gets through, and he's expecting the law to step in on what's safe to use. As for the future, he's not sweating about AI taking over his job. He's more into how it can make work less of a grind for everyone.

What'd you think of today's edition?

Login or Subscribe to participate in polls.

What are MOST interested in learning about AI?

What stories or resources will be most interesting for you to hear about?

Login or Subscribe to participate in polls.

Join the conversation

or to participate.