| |
| Hello, and happy Friday! Welcome to Inside AI. Here are the top stories for today: - The European Union is looking into alleged anticompetitive practices in the AI chip market.
- Apple is expanding its generative AI research and workforce in the U.K.
- Meta used public Facebook and Instagram posts to train its new virtual assistant.
If you find this newsletter valuable, please consider sharing it with your friends and colleagues. Thank you! Beth
p/beth-duckett | |
| 1 | | The European Union and French authorities are investigating potential anticompetitive practices in the AI chip market, primarily led by Nvidia. On Tuesday, France's competition authority raided Nvidia's French offices on suspicion of anticompetitive practices in the graphics cards industry, according to the Wall Street Journal. More: - The European Commission is now gathering opinions on potential abuses in the GPU sector for AI and gaming to determine if future intervention is needed.
- Likewise, French authorities have been interviewing industry insiders about Nvidia's role in AI chips, pricing strategies, chip shortages, and their price impacts.
- The French authority conducted this week's raid as part of its scrutiny over cloud technology, clarifying that raids don't assume a legal violation.
Zoom out: - Nvidia holds an over 80% share in the GPU market. The company has thrived in the AI boom with high demand for its AI chips, boosting its revenue and market cap to over $1T.
|     | |
| 2 | | Apple is actively researching generative AI technology and plans to increase its AI workforce in the U.K., according to CEO Tim Cook. Cook also confirmed Apple's research into a ChatGPT-like service. More: - In addition, Cook highlighted the role of AI in Apple products, including features like Fall and Crash Detection, atrial fibrillation detection, ECG functionality, and predictive autocorrect.
- AI is "literally everywhere on our products," he said, "and of course, we're also researching generative AI as well, so yes, we have a lot going on."
- Apple has ramped up its spending on AI technologies to millions of dollars per day, according to a report from The Information.
- Apple's advanced LLM, internally named Ajax GPT, is reported to surpass OpenAI's GPT-3.5 in power, boasting over 200 billion parameters.
- Apple is actively developing a chatbot for AppleCare interactions and another to streamline complex tasks with Siri.
Question: Which of Apple's AI features are you most excited about or find the most valuable? Share your thoughts and join the conversation here. | | | |
| |
|
| 3 | | Meta used public Facebook and Instagram posts to train its new virtual assistant, Meta AI. The company excluded private posts shared with family and friends to respect users' privacy, according to Meta's President of Global Affairs, Nick Clegg. More: - Meta's AI assistant, unveiled at Wednesday's Connect event, is being integrated into WhatsApp, Facebook Messenger, and Instagram.
- Powered by Meta's open-source large language model Llama 2, the chatbot can generate text and images, access the internet, and answer user queries.
- While some of the AI assistant's training comes from public Facebook and Instagram posts, Clegg told Reuters that Meta tried to filter out datasets with "a heavy preponderance of personal information."
- The "vast majority" of the data was publicly available, he noted.
Zoom out: - Meanwhile, companies like Meta, OpenAI, and Google have faced criticism for using internet data without consent to train their AI models.
- In July, comedian/author Sarah Silverman and authors Christopher Golden and Richard Kadrey sued Meta and OpenAI, alleging copyright infringement of their online works.
| | | |
| 4 | | Google's upcoming smartphones, the Pixel 8 and Pixel 8 Pro, will have AI-powered camera features. These include "Photo Unblur" for post-capture picture unblurring and "Magic Eraser" for removing unwanted objects from photos. More: - Google will announce its latest Pixel devices during its hardware event on Oct. 4.
- A spec sheet hints at both smartphones running on Google's new Tensor G3 chipset with 8GB of RAM.
- While the AI camera features were included in the Pixel 7 phones, the new Tensor G3 chip could boost their speed and accuracy.
Zoom out: | | | |
| |
|
| 5 | | Organizations are most concerned about inaccuracies and cybersecurity breaches in generative AI systems, according to McKinsey's Global Survey on AI. 
What the numbers show: Organizations see inaccuracy as the primary risk in generative AI, with cybersecurity ranking second. Only 21% of AI adopters have established policies for gen AI use at work, and 32% are working to mitigate inaccuracy compared to 38% for cybersecurity. Most companies aren't addressing the political, environmental, or national security risks of genAI, deeming them beyond their scope. Why it matters: Alexander Sukharevsky, senior partner and global leader at QuantumBlack AI by McKinsey, says that while there is broad awareness of generative AI's risks, pervasive anxiety and fear can hinder effective risk management. He suggests that many of these risks can be mitigated through changes in a company's technology architecture aligned with its policies. The most constructive approach involves experimenting with generative AI within a structured process to pinpoint and address the broader risks. | | | |
| 6 | | Weekly roundup — The top AI stories you might have missed this week: - Former Apple designer Jony Ive and OpenAI CEO Sam Altman are said to be exploring an AI-powered smartphone with over $1B in backing from Softbank CEO Masayoshi Son, per the Financial Times. Ive's design agency, LoveFrom, would develop the "iPhone of artificial intelligence" with the goal of creating "a more natural and intuitive user experience for interacting with AI."
- Amazon plans to invest up to $4B in Anthropic, an AI startup founded by former OpenAI staffers. Amazon's initial investment will be $1.25B, with the chance for an additional $2.75B. Amazon will gain a minority stake in Anthropic, known for its AI-powered chatbot Claude 2.
- Meta unveiled a range of new AI services at its Connect event. Meta's new AI assistant and chatbot generates both text and images on WhatsApp, Messenger, and Instagram. It's also rolling out 28 AI characters in its messaging apps, including ones based on celebrities like Charli D'Amelio, MrBeast, and Snoop Dogg.
- OpenAI is in talks for a potential share sale that could value the company at up to $90B, significantly higher than its nearly $30B valuation earlier this year. According to the Wall Street Journal, the deal is expected to allow OpenAI employees to sell their shares rather than the company issuing new ones.
- Amazon Bedrock, the company's suite of generative AI tools, is now generally available for businesses to build their own AI applications. The toolkit features foundation models for text and images from AI21 Labs, Anthropic, Stability AI, and other third parties, along with Amazon's own Titan large language models.
| | | |
| 7 | | "I believe that we as humans want human content … I'm focusing on human-generated content while other people are focusing on AI." Justin, founder of Quda, is not your average technologist. With a degree in theology from Cambridge and a wide area of expertise, Justin learned Bubble and used the no-code tool to build in public over the past few years. This fascinating conversation covers no code, AI, social media, and work-life balance for founders. In a selected Q&A summary below, Justin talks about his optimism around human-generated content and the ancient format of question and answer being the atom of learning. 
| | | |
| 8 | | Quick Hits: - No-code development is here to stay, according to industry experts. Gartner predicts 80% of technology products will be built by those who are not IT professionals by 2024.*
- Jasper AI, a generative AI startup, has reduced the values of internal common shares by 20%, lowering employees' stock values and suggesting a potential slowdown in its AI writing tool growth. The company, based in Austin, has also hired former Dropbox president Timothy Young to be its new CEO.
- At this week's Code Conference, Bumble CEO Whitney Wolfe Herd discussed how the dating app plans to leverage AI to improve matchmaking, reducing the time it takes for users to meet in person.
- 56% of companies surveyed said they are not using generative AI tools, such as ChatGPT, in their HR endeavors, according to Littler's AI in the Workplace Survey Report.
- NEXT WEEK in LA, redefine development at DockerCon! Use DCHALF for 50% off in-person registration, or participate online for free. Come learn with us!*
*This is a sponsored post. | | | |
| Upcoming Events | | OCT
16 | | Meet Our Fund 5, Building and Investing in AI | | | | | | * This is a sponsored event | | | |
Term of the Day Minimum viable product (MVP): A minimum viable product (MVP) is an early functioning model of a new product that can be presented to early adopters and investors. Question of the Week Do you share details about your salary with your coworkers? Join the conversation |
INSIDE AI LEADERBOARD (7 DAYS) |
 | AI and technology writer | | Beth is an editor and writer of Inside's AI and Tech newsletters, with a career spanning journalism since 2007. She has written for publications including The Arizona Republic and USA Today and authored a book on the solar industry in 2016. Reach her at Beth.Duckett@yahoo.com. | | This newsletter was edited by Aaron Crutchfield | |
| |
 | Get your own DevSquad, a high-performing product team that will build your SaaS in 3 to 6 months. | |
|
 | See the power of the Wiz platform. | |
|
| |
Comments
Post a Comment