Today:

Hey folks!

In today's newsletter, we're thrilled to continue our exploration of the revolutionary Autogen Studio, building on the foundation laid in our previous video titled "AutoGen Tutorial 🔥 How to Build POWERFUL AI Agents". This serves as Part 2 of our journey into the incredible capabilities of Autogen Studio, where we delve deeper into its features and functionalities.

The latest Autogen release simplifies the setup process and introduces new features like skills, allowing you to program by talking to AI agents. The video offers a step-by-step guide on how to install and use Autogen Studio, making it accessible even to those without extensive programming experience.

WATCH THE VIDEO ON YOUTUBE

AI Will Transform the Global Economy. Let’s Make Sure It Benefits Humanity.  

AI's rapid growth could revolutionize productivity and global growth but also poses risks like job loss and increased inequality. Nearly 40% of global jobs are at risk, impacting both low and high-skill sectors. Advanced economies might face more challenges but also have better opportunities to use AI. 

Meanwhile, emerging markets and developing countries, less exposed to AI, lack infrastructure to fully benefit from it, potentially widening global inequality. Within countries, AI could increase income gaps, benefiting those who can utilize it. 

To address these challenges, comprehensive social safety nets, retraining programs, and inclusive policies are essential. The IMF's AI Preparedness Index helps countries assess their readiness, highlighting the need for developed nations to innovate responsibly and for developing nations to build a digital foundation.

IMF

Microsoft’s new Copilot Pro brings AI-powered Office features to the rest of us 

Microsoft just dropped Copilot Pro for the everyday user. Before, it was just for the business crowd. For 20 bucks a month, you can now jazz up your Office apps like Word and PowerPoint with some cool AI tricks. Think creating full-blown presentations from a simple chat prompt or reworking paragraphs in Word. Plus, if you're already paying for Microsoft 365, this is just an extra 20 per person each month.

On top of making your Office stuff smarter, Copilot Pro hooks you up with the latest AI models from OpenAI, better image creation tools, and lets you build your own AI buddy. For those not on Microsoft 365, grabbing Copilot Pro alone gives you VIP access to faster AI performance and new features like better image creation and custom AI creation.

For the business folks, Microsoft's also making their Copilot for Microsoft 365 more accessible, dropping the minimum seat requirement and offering it for $30 per user each month.

THE VERGE

Elon’s Tesla robot is sort of ‘ok’ at folding laundry in pre-scripted demo

Elon Musk's Tesla showed a video of their robot, Optimus, folding a T-shirt. It's kinda impressive, but don't get too excited. First off, it's slow. Also, it's not thinking for itself – it's just following a set script, like a fancy puppet. Musk says one day it'll do this stuff on its own, but right now, it's like watching a robot play-acting.

The robot's good at moving its parts, but that doesn't mean it's smart. Musk hinted at big future improvements, but considering where tech's at, that seems like a stretch. So, for now, keep folding your own laundry – Optimus isn't taking over that chore anytime soon.

TECHCRUNCH

China's military and government acquire Nvidia chips despite US ban

Despite the U.S. ban, Chinese military bodies, AI research institutes, and universities have been buying small batches of Nvidia chips from lesser-known Chinese suppliers. This situation underscores the challenge for the U.S. in fully blocking China's access to these advanced chips, which are crucial for AI and military computing developments. 

The demand for Nvidia chips remains high in China, where alternatives are still emerging. The chips, including the banned A100 and H100 models, are being used for AI tasks, although the quantities purchased are not large enough to build sophisticated AI systems from scratch. 

Nvidia and approved retailers are not the suppliers, suggesting an underground market for these chips in China. Nvidia asserts its compliance with export control laws. The U.S. is working to close loopholes in the export ban, aiming to hinder China's AI progress. 

REUTERS

Baidu Sinks Most Since 2022 Despite Denying Links to PLA AI

Baidu's stock took a nosedive, the biggest since 2022, after a report linked its AI, Ernie, to China's military research. This news freaked out investors, scared of possible U.S. sanctions against Chinese tech due to military connections. Baidu, a big name in AI in China, saw its shares drop over 10% on Jan 15 in Hong Kong. 

The panic started with a South China Morning Post article saying a Chinese military university tested AI on Baidu’s Ernie. Baidu denied any ties with the military. The fear is that the U.S. might hit Chinese firms with sanctions. 

Adding to the worry, Baidu’s facing a rough patch in China’s economy and is struggling to make big bucks from Ernie, which is still new. Investors are jittery, selling off shares, wary of U.S.-China tensions and unsure if the military link is even true.

BLOOMBERG

MEET THE TINY STARTUP AIMING TO TAKE ON APPLE AND META WITH AI-POWERED SMART GLASSES

Brilliant Labs, a gutsy three-person startup, is challenging tech titans like Apple and Meta in the augmented reality (AR) glasses game. Based in Singapore's industrial zone, they've teamed up with Moveon Technologies to craft high-tech optics. Their CEO, Bobak Tavangar, envisions AR glasses with AI smarts, aiming for an open-source platform to attract developers.

They're set to launch smart glasses with natural language processing soon, eyeing more features like visual cues and adjustable views. Despite the rocky history of AR, Brilliant Labs is playing it cool, focusing on profitability over quick wins.

Their first product, the Monocle, is an AR lens for regular glasses. It's voice-operated, loaded with cool features like ChatGPT and sells for $299. It's a hit with over 3,000 units sold. Backing them are big names like Siri co-founder Adam Cheyer, impressed by their vision-enhancing approach. Despite not revealing their valuation, they've pocketed $6 million from top investors.

FORBES

🧠RESEARCH

PhotoMaker: Customizing Realistic Human Photos via Stacked ID Embedding

The "PhotoMaker Style" created by TencentARC, is an innovative tool that allows users to customize realistic human photos via Stacked ID Embedding. It's easy to use and offers a creative way to personalize images. This tool is particularly useful for those looking to add an artistic flair to their digital images. 

facebook/seamless-expressive 

The "SeamlessExpressive'' model by Facebook combines two innovative modules. The Prosody UnitY2 module translates speech to units with awareness of speech prosody, like speech rate and pauses. The PRETSSEL module converts units to speech, maintaining cross-lingual expressivity, like voice style. This tech allows for expressive, multilingual speech translation, featuring different voice styles and preserving the expressiveness across languages. 

MAGNeT

MAGNeT, part of Facebook's collection on Hugging Face, introduces a novel approach to audio generation using a single non-autoregressive transformer. This technology focuses on creating audio content efficiently, and the collection includes various models that handle text-to-audio conversion. The models are differentiated based on their sizes and the length of audio they can generate, offering diverse capabilities in audio synthesis. This innovation marks a significant step in audio technology, showcasing the potential of transformers in multimedia applications. 

🛠️TOP TOOLS

PromptLayer - A platform for managing, evaluating, and logging prompts in AI systems. Features include version control, operational efficiency, and collaboration tools for teams. 

HoneyDo - voice-powered grocery list app that simplifies shopping by converting spoken words into organized lists. It includes features like saving original voice memos, shared lists with real-time iCloud updates, and a family sharing option. 

Athina - monitoring and evaluating Large Language Models (LLMs) in production. It helps in detecting and fixing hallucinations and quality issues in AI outputs. Features include monitoring LLM usage, debugging outputs, and gaining conversational insights.

Antispace - digital command center that integrates various software applications like Mail, Calendar, Notes, Todo, Slack, and GitHub into a single interface, facilitating seamless navigation and task management. 

Trolly.ai - AI-powered platform designed to create professional SEO articles quickly. It features a text editor, supports over 20 languages, and provides tools like grammar checkers, in-article images, and team access.

📲SOCIAL MEDIA

🗞️MORE NEWS

How OpenAI is approaching 2024 worldwide elections

OpenAI is gearing up for the 2024 global elections, focusing on preventing misuse of AI, like misleading deepfakes or chatbots faking as candidates. They're ramping up efforts to keep the election legit, like tweaking ChatGPT to avoid political campaigning, ensuring bots don't pretend to be real folks or institutions, and nixing apps that discourage voting. They're also making moves to help people spot AI-created images and teaming up with NASS in the U.S. to link folks to legit voting info. They're all about keeping it real and fair for the elections. OPENAI

Spot Technologies, now with $2M, will see AI security tech go into Mexico Walmarts

Spot Technologies, an AI startup based in El Salvador with operations in Chile, has secured $2 million in funding. The company is developing cloud technology that transforms cameras in retail and logistics locations into an intelligent system for behavior analysis and security. Their flagship product, VisionX, utilizes deep learning and computer vision to analyze consumer and theft behaviors, offering features like gender and age analysis, people counting, and more. Spot Technologies plans to use the funding to further develop VisionX 2.0, which will include features for detecting thefts and analyzing shopping patterns. Walmart is one of their major customers, having deployed VisionX in its stores and distribution centers in Chile. TECHCRUNCH

Nikon develops AI system that can detect cows are about to give birth

Nikon has developed an AI system that can detect when cows are about to give birth by analyzing their movements with cameras installed on farms. This technology is aimed at increasing efficiency and reducing the burden on farmers who need to conduct regular checks on pregnant cows leading up to giving birth. The system is estimated to cost approximately ¥900,000 ($6,200) per year for a farm with 100 cows and uses a dedicated smartphone application to alert farmers when a calf is due. The AI system has been developed with data collected since the fall of 2021 and has undergone proof-of-concept experiments on farms in Kumamoto Prefecture, Japan. THE JAPAN TIMES

ChatGPT for Self-Diagnosis: AI Is Changing the Way We Answer Our Own Health Questions

AI-powered chatbot ChatGPT is changing the way people self-diagnose health issues. It offers a more personalized search for health information, potentially saving time in a healthcare system with long wait times and communication gaps. However, it comes with risks, including the possibility of providing incorrect information. Doctors see ChatGPT as a complementary diagnostic tool, not a replacement for medical professionals. Patients can benefit from using ChatGPT to frame questions for doctors and gather general information, but they should exercise caution and not solely rely on it for medical advice. CNET

Amazon Is Selling Products With AI-Generated Names Like "I Cannot Fulfill This Request It Goes Against OpenAI Use Policy"

Amazon is selling products with AI-generated names, such as a dresser named "I'm sorry but I cannot fulfill this request it goes against OpenAI use policy." These product listings appear to be generated using OpenAI's ChatGPT without proper proofreading, likely in an attempt to improve search engine optimization. Multiple products on Amazon now feature similar AI-generated notices in their descriptions. This raises concerns about the quality control of products listed on Amazon and highlights the potential misuse of AI for creating product descriptions without human oversight. Amazon has removed some of these listings and stated its commitment to enhancing its systems for accurate product listings. FUTURISM

This AI-enabled e-bike uses ChatGPT to give you a smarter ride

Urtopia, a tech-savvy e-bike company, introduced an AI-enabled e-bike at CES 2024, integrating ChatGPT for a smarter ride. This "Jarvis" of bikes connects via LTE and Bluetooth, offering voice commands without requiring a smartphone. The bike acts as a coach and guide, suggesting routes based on your fitness condition, while an upcoming smart ring syncs health data for a tailored biking experience and doubles as a bike security key. The Urtopia Carbon 1 Pro is available for pre-order at $2,499. ZDNET

After AI’s summer: What’s next for artificial intelligence?

In 2023, AI witnessed remarkable advancements, particularly in Large Language Models (LLMs) and chatbots, as well as in image, video, and voice generation. People increasingly embraced AI in their daily work, fueling predictions of future innovations, including home robots and artificial general intelligence (AGI) within a decade. Physical robotics, driven by LLMs and image recognition, are poised to enhance human-robot interactions. However, challenges such as data bias, energy consumption, copyright issues, and existential threats need addressing. Despite potential hurdles, AI's momentum is expected to continue in 2024, with rapid technological improvements and potential life-changing innovations on the horizon. VENTUREBEAT