- WKND AI
- Posts
- AI News Roundup
I hope you had a fantastic Labor Day weekend!
For those who, like me, took some time off to travel or simply relax during the holiday, diving back into the ever-evolving world of AI might feel a bit overwhelming. But fear not! This week, we're doing things a bit differently.
If you've ever felt behind on the latest in AI, this is your golden ticket to catch up. The month of August was a busy one for AI news, so we've curated this month’s top news to get you back on track no matter how busy your month was.
But that's not all. We're also introducing some exciting new segments to our newsletter:
📸 Real or AI? - Can you spot the difference between reality versus AI-generated content? Take our poll and see for yourself. We'll reveal the results next week, so stay tuned!
🛠️ AI Tools Of The Week - Discover AI tools that have caught our attention. Whether you're in the industry or just an enthusiast, these tools promise to add value. And hey, if you're an AI tool creator or know one, this could be a potential spotlight for you in the future!
🎨 AI Images Of The Week - Whether they're crafted by our team on Midjourney or submitted by one of you talented folks, they're sure to inspire.
As we continue to evolve and incorporate more diverse content, your feedback is invaluable. Let us know what you think, and if there's anything else you'd love to see in our future editions.
AI News Roundup
Researchers at UCSF and UC Berkeley have developed an AI-powered brain implant that translates brain signals into speech and facial expressions. A paralyzed woman, unable to speak due to a stroke, communicated and conveyed emotions using a digital avatar. The breakthrough utilizes advanced AI to decode natural speech, offering hope for those with communication disabilities. The study, led by UCSF neurosurgeon Edward Chang, was published in the journal Nature.
Tesla launches its supercomputer, one of the world's most powerful, employing 10,000 Nvidia H100 GPUs. Aimed at AI applications, it will also handle high-performance computing workloads. This advancement will expedite the training of Tesla's full self-driving (FSD) technology. Tesla's investment in AI training is evident, with plans to spend over $4 billion in the next two years on FSD training.
Amazon introduces AI-generated customer review highlights, offering a concise overview of product features and sentiments from reviews. This feature, available to a subset of mobile users in the U.S., provides insights on product attributes like "ease of use." Amazon's commitment to authentic reviews remains, with strict measures against fake reviews and a focus on verified purchases.
AI startup Hugging Face has garnered $235 million in a Series D funding round, elevating its valuation to $4.5 billion. With participation from tech giants like Google, Amazon, and Nvidia, the company offers data science tools, AI code repositories, and an enterprise version of its hub. The funding will bolster efforts in research, enterprise, and startups.
OpenAI, the company behind ChatGPT, has filed a trademark application for GPT-5 with the US Patent and Trademark Office (USPTO). The application hints at OpenAI's development of a voice assistant capable of live speech translation and generating its own AI models. While the current ChatGPT version operates on the GPT-4 model and is text-based, the GPT-5 iteration might introduce voice capabilities, similar to Alexa. The trademark also suggests potential features like machine learning dataset sharing and neural network development.
OpenAI has rolled out the 'custom instructions' feature for all ChatGPT users, granting more control over the AI's responses. Initially launched in July for ChatGPT Plus subscribers, this feature lets users set preferences for the AI's replies, eliminating the need to repeatedly input the same instructions. Users can now specify response length, tone, and other criteria. Previously exclusive to paying subscribers, it's now available for both free and Plus users, including on mobile platforms.
OpenAI is introducing several enhancements to the ChatGPT experience, which include:
Prompt examples at the start of new chats for inspiration.
Suggested replies to enrich conversations.
GPT-4 as the default model for Plus users, retaining user preferences.
Capability to upload multiple files for data analysis and insights, available with the Code Interpreter beta for Plus members.
Extended login sessions and a friendlier login page.
Handy keyboard shortcuts for efficient interactions.
OpenAI unveils ChatGPT Enterprise, offering enhanced security, unlimited GPT-4 access, longer input processing, and advanced data analysis. Aimed at elevating work productivity, it's adopted by over 80% of Fortune 500 companies. Features include data encryption, SOC 2 compliance, and a 32k token context. Partnerships with industry leaders like Block and Canva highlight its transformative potential in business operations.
The New York Times (NYT) may be gearing up to sue OpenAI after updating its terms of service to prevent AI companies from using its content. A lawsuit could lead to the destruction of ChatGPT's dataset and hefty fines for OpenAI. The Times' concerns include potential competition from ChatGPT, which might use NYT's content to answer questions based on the paper's original reporting.
Despite ChatGPT's prominence in tech, Pew Research indicates only 18% of Americans have used it. Usage varies by demographics, with 30-40% among men, ages 18-29, and college-educated. Most use it for entertainment or learning rather than work. Among those aware of ChatGPT, 19% believe it will majorly impact their job, while 27% foresee no impact. Concerns about AI's role in daily life have risen, with 47% more concerned than excited.
President Joe Biden has signed an executive order prohibiting certain new U.S. investments in China, specifically in sensitive tech sectors like semiconductors and AI. This move aims to prevent American resources from aiding China's technological advancements that could bolster its military and threaten U.S. national security. The order targets various investment forms, including venture capital and joint ventures. While it addresses immediate national security concerns, it has sparked debates on its broader implications for the global economic landscape.
The XQ-58A Valkyrie drone successfully completed a test at Eglin Air Force Base, showcasing its AI capabilities in air combat scenarios. Developed by the Air Force Research Laboratory (AFRL), the drone's AI algorithms were trained through millions of hours in high-fidelity simulations. The test demonstrated the AI's ability to solve a tactically relevant 'challenge problem' during flight. This achievement builds upon previous autonomous aircraft programs like Skyborg and the Autonomous Aircraft Experimentation (AAx). The Air Force emphasizes the importance of AI in future warfare, with Brig. Gen. Scott Cain stating that AI will be a "critical element to future warfighting".
Zoom's recent terms of service update reveals its decision to use some customer data for training its AI models. Effective from July 27, the update allows Zoom to utilize certain "service-generated data" for AI training, excluding audio, video, or chat content without customer consent. This move aligns with Zoom's AI ambitions and has sparked discussions on the ethics of using individual data for AI, even if anonymized.
A recent study reveals that artificial intelligence can determine which keys are being pressed on a laptop just by listening to the sound of the typing. With over 90% accuracy, this AI system can identify keystrokes from sound recordings. As video conferencing tools like Zoom become more prevalent, the potential for cyber-attacks based on such acoustic cues rises. The research emphasizes the importance of public debates on AI governance and the need for vigilance in digital security.
Lil Miquela, a virtual influencer, boasts an eight-figure net worth and charges over $10,000 per Instagram post. With 3.6 million TikTok followers and 2.7 million on Instagram, she's a digital sensation. Yet, she's not human. Created by the American AI company Brud, Miquela has collaborated with global brands like Dior, Chanel, and Alexander McQueen. She even secured a multi-million-dollar deal with Samsung in 2019. As AI influencers rise in popularity, they offer brands unparalleled flexibility, reshaping the advertising landscape.
Real or AI?
Image 1
Can you tell the difference between AI and reality?Select the image you think is real. |
Image 2
AI Tools Of The Week
BeFake: AI-powered social media for digital self-expression. Transform text prompts into captivating visuals beyond physical limitations.
Wirestock: Effortless platform to sell photos, videos, and AI art. No keywords or captions needed, just upload and sell.
InVideo: Simplified video creation. Convert text to video, use templates, and enhance brand exposure.
AI Images Of The Week
Image generated by Midjourney. Try out the prompt below.
Taylor Swift dressed as black metal singing in school cafeteria, full body, dynamic lighting, unreal engine, Hasselblad X2D
Image generated by Midjourney. Try out the prompt below.
ghost marshmallows, happy, pastel colors, in style of adventure time, intricate detail, concept artHow'd you like this newsletter?Your feedback helps us make cooler emails for you! |
Reviews Of The Week
To support this newsletter, tell a friend to subscribe here!
