Today:

GPT store pics leaked + Can GPTs pull in $20,000+ per month? Here's how to start building.

Everybody's buzzin' about OpenAI's GPT store, rumored to launch any day now. There's this leaked screenshot floating around, showing what the store might look like. It's got all these GPTs listed, like this dude Danis Koves topping the charts with his GPT-5. Turns out, Koves is a whiz at Google ranking. 

The store's gonna have categories like writing and programming to help you find what you need. Here’s a quick dive on how to set up your profile, making dough with GPTs, and showcases some potential big hitters, like PDF and chatbase apps. 

WATCH THE VIDEO ON YOUTUBE

Google faces $1.67 bln damages demand at AI-related patent trial

Google's in a big legal showdown in Boston, facing accusations they swiped tech from Joseph Bates, a computer whiz from Massachusetts. His company, Singular Computing, says Google used his smarts to beef up AI in stuff like Google Search and Gmail.

They're asking for a whopping $7 billion, which would smash the record for patent lawsuits in the U.S. Google's brushing it off, claiming their tech was cooked up in-house over years. The trial's gonna take about two to three weeks. Singular's beef is about Google's Tensor Processing Units, fancy circuits they say rip off Bates' inventions. Google argues their gear is different and Bates' patents don't hold water. Meanwhile, another court's deciding if those patents should even exist. 

REUTERS

EU checking if Microsoft’s OpenAI investment falls under merger rules

The EU is scrutinizing Microsoft's investment in OpenAI to see if it falls under merger rules. This follows a dramatic episode at OpenAI, where founder Sam Altman was ousted then rehired, with Microsoft backing him. This event led to a reshuffled OpenAI board, with Microsoft gaining a board observer role. Microsoft owns 49% of OpenAI's for-profit arm, raising regulatory eyebrows.

The UK and Germany have already probed this partnership, with the EU now joining in. They're inviting input on competition in AI and virtual worlds, focusing on how these emerging markets are influenced by big tech partnerships. The EU's keen on understanding market dynamics in these areas, especially with Microsoft and OpenAI's growing clout. The outcome could influence how digital competition laws, like the EU's Digital Markets Act, apply to rapidly evolving tech sectors like AI.

TECHCRUNCH

Rabbit’s Little Walkie-Talkie Learns Tasks That Stump Siri and Alexa

Jesse Lyu, founder of Rabbit Inc., introduces the R1, a nifty AI gadget, at CES 2024. Unlike smartphones packed with apps, the R1 is a small, red-orange device that uses voice commands and AI to perform tasks. It's designed with a touchscreen, rotating camera ("Rabbit Eye"), and a simple push-to-talk button. Priced at $199, it's not a smartphone replacement but more like a personal assistant for tasks like ordering Ubers or making dinner reservations.

The R1 operates without apps or smartphone pairing. Instead, it uses Rabbit OS and large language models, like ChatGPT, to understand and execute tasks. The device is privacy-conscious, not storing user credentials or constantly listening. A unique feature is "teach mode," allowing users to train it for specific tasks. Lyu envisions the R1 as a seamless

WIRED

Unlocking a new era for scientific discovery with AI: How Microsoft’s AI screened over 32 million candidates to find a better battery

Microsoft's Quantum team, using AI, has screened over 32 million potential materials, finding a breakthrough material for better batteries. This AI-powered discovery, part of the Azure Quantum Elements initiative, showcases the power of AI, cloud computing, and eventually, quantum computers in accelerating scientific discovery. They collaborated with Pacific Northwest National Laboratory (PNNL) to identify and synthesize this novel material, reducing lithium use by about 70%, making it a more sustainable energy-storage solution.

This AI-driven approach speeds up research from years to days. It's a major step in using AI for scientific research, bringing new possibilities in chemistry and materials science. This breakthrough is just the start of a new era in scientific discovery, offering significant environmental, safety, and economic benefits.

MICROSOFT

SandboxAQ Buys Good Chemistry, Melding AI and Quantum Tech for Drug Discovery

SandboxAQ, an AI and quantum tech firm, snagged Good Chemistry for about $75 million. This big move aims to speed up making new drugs and materials, using the smarts of quantum computing. Jack Hidary, the head honcho at SandboxAQ, says this deal will help them serve more customers and bring in some top-notch quantum talent.

Good Chemistry, from Vancouver, knows its stuff in predicting and simulating chemical properties, a skillset that SandboxAQ wants to beef up. They hope this will make creating new drugs and materials faster and cheaper – a real game-changer. This is SandboxAQ's second buy, after getting a cybersecurity company last year. The tech world is buzzing, seeing this as a smart match in the hot AI and quantum scene.

THE WALL STREET JOURNAL

Amazon’s Alexa gets new generative AI-powered experiences

Amazon's Alexa is stepping up its game with new AI-powered skills. At CES, Amazon announced three fresh Alexa experiences from Character.AI, Splash, and Volley, now available in the Alexa Skill Store.

Character.AI lets users chat with fictional personas or celeb and historical figure imitations like Elon Musk or Shakespeare. You can even role-play in adventures, like aiding Mario to rescue Princess Peach. Splash offers a cool Alexa Skill to create personalized songs in various genres, which you can download. Volley's bringing a classic "20 Questions" game, but with a generative AI twist.

Alexa itself got smarter, with an AI model for more human-like emotions and opinions. Amazon's also introducing fun, kid-friendly learning features for Amazon Kids+ subscribers. And it's not just Alexa – Amazon's CES news includes Fire TV on Panasonic smart TVs in 2024, Matter Casting on Fire TV and Echo Show 15, and updates on their self-driving robotaxi Zoox.

TECHCRUNCH

🧠RESEARCH

Mixtral of Experts

The paper introduces Mixtral 8x7B, an advanced language model with unique features. It's built like Mistral 7B but has 8 special blocks in each layer, choosing two per token for processing. Despite its massive 47B parameter size, it only actively uses 13B during tasks. Excelling in areas like math and coding, Mixtral beats major competitors like GPT-3.5 and Llama 2 70B. A special version, Mixtral 8x7B – Instruct, tops other models in human benchmarks. Both versions are available under the Apache 2.0 license, with codes and details online.

MoE-Mamba: Efficient Selective State Space Models with Mixture of Experts

This paper introduces MoE-Mamba, a model combining State Space Models (SSMs) and Mixture of Experts (MoE) to outdo standard Transformers in sequential tasks. MoE-Mamba trumps both Mamba and Transformer-MoE models, achieving similar results as Mamba with over twice the efficiency in training, while maintaining Mamba's superior performance during use.

DocGraphLM: Documental Graph Language Model for Information Extraction

DocGraphLM is a new tool by JPMorgan's AI team. It’s a cutting-edge approach combining language models with graph technology for enhanced information extraction from visually complex documents. It introduces a dual-part encoder for document representation and an innovative link prediction method for reconstructing document graphs. This model improves direction and distance prediction between nodes, focusing on neighborhood restoration. Tests on three top datasets reveal significant improvements in information extraction and question answering tasks, with faster learning due to graph feature integration.

🛠️TOP TOOLS

Emma AI - AI-powered assistant tool that enables quick and easy creation of custom chatbots using OpenAI's GPT-3.5 technology. It integrates with various data sources and applications, offering personalized, efficient assistance across different teams and workflows. 

Aragon AI - Turns selfies into professional AI-generated headshots. It requires only a few photos to create varied headshots with different backgrounds and styles. 

McAnswers - Simplifies the coding journey by leveraging ChatGPT 4.0 technology. It provides solutions to common coding errors and bugs across various programming languages, offering detailed explanations and fixes. 

Invstr -  All-in-one investment research platform. It uses AI to analyze financial reports, offering quick summaries, risk analysis, and insights from earnings calls. 

Castmagic - Amplify audio content creation using AI. It converts media files into diverse content assets like transcripts, social media posts, email newsletters, and more. 

📲SOCIAL MEDIA

🗞️MORE NEWS

OpenAI and journalism

OpenAI and The New York Times were discussing a partnership until the Times sued OpenAI unexpectedly on December 27. OpenAI claims the Times didn't help address content reuse concerns and may have manipulated tests to provoke specific responses from ChatGPT. OpenAI is working to make its systems more robust against such misuse. OPENAI

X promises peer-to-peer payments, AI advances in 2024

Elon Musk's X, previously Twitter, plans to transform into an "everything app" with features like peer-to-peer payments and AI enhancements. Despite Musk's past delays, X is progressing, already licensed in 14 states for payments. Users will soon send money and possibly get high-yield accounts. X will also focus on creator support, online shopping, and ad safety, despite past ad placement issues. Musk's AI company, xAI, will enhance user experience and advertising, including features to broaden viewpoints. Last year, X introduced several new services and reported significant user engagement. TECHCRUNCH

Walmart is betting that AI can help you shop faster — and maybe do your shopping for you

Walmart is innovating shopping with AI. Their InHome service, costing $20 monthly, might soon auto-order and deliver groceries, refilling items like milk magically. This AI model learns your habits and Walmart's consumer data. Unlike Amazon's Dash Replenishment for specific items, Walmart's AI aims for responsive, personalized grocery restocking. At CES 2024, Walmart showcased a generative AI search engine for broader queries and a Sam’s Club device to auto-check carts, reducing checkout lines. Walmart's vision is faster, easier shopping, including drone deliveries and AI-assisted shopping, sometimes doing the shopping for you. THE VERGE

X's former head of trust and safety Ella Irwin joins Stability AI

Ella Irwin, ex-head of trust and safety at Elon Musk's X, joins Stability AI as Senior VP of Integrity. She'll tackle child safety, intellectual property, and disinformation issues, arriving amid AI's rapid growth and ethical concerns. Her move follows a challenging stint at X, where she managed controversies and left over a misalignment of principles. Stability AI, credited with Stable Diffusion, focuses on AI's safe use, including initiatives against disinformation and promoting good practices in AI development. Irwin's role is pivotal as AI's influence and potential misuse in areas like elections and child safety become increasingly significant. NBC NEWS

PhotoRoom, the AI photo editing app, is raising $50M-$60M at a $500M-$600M valuation

PhotoRoom, a Paris-based AI photo editing app, is raising $50M-$60M at a $500M-$600M valuation, making waves in AI and mobile image manipulation. Targeting e-commerce, media, and various users, PhotoRoom's AI app and API offer quality, easy-to-use photo editing. It's known for powering a Barbie personalized poster site and attracting significant investment interest despite a tight funding environment. Balderton, a previous investor, might lead this round, alongside other existing and new investors. PhotoRoom's success reflects the growing importance of cameras in smartphones and its unique B2B focus. It offers a freemium model with varying pricing tiers and has gained popularity, ranking high in app stores. The app's downloads have surpassed 100 million, with an anticipated annual revenue of $50 million. TECHCRUNCH

Pennsylvania government workers will start using ChatGPT in test program

Pennsylvania is starting a pilot program using OpenAI's ChatGPT for administrative tasks in government work. This initiative, the first of its kind for ChatGPT Enterprise in a state, aims to assist employees in the Office of Administration with tasks like writing, policy language updating, job description drafting, and coding. The program will initially involve a select group of employees, with plans to extend it based on feedback and potentially add 100 more licenses. THE VERGE

The industries where AI-powered robots will arrive next, according to Nvidia

Nvidia's VP of Embedded and Edge Computing, Deepu Talla, predicts that AI-powered robots will soon revolutionize construction, retail, healthcare, and agriculture. These industries follow logistics, warehousing, and manufacturing, which already benefit from robotic efficiencies. Talla highlights that robots, increasingly used for tasks like labor shortages and cost reduction, are becoming smarter with AI advancements. Generative AI is notably changing robot development by accelerating the transition from concept to deployment. These robots often work in unpredictable environments, especially with human presence. Nvidia's generative AI model, Picasso, helps in simulating realistic conditions like rain or rust, vital for testing and training robots. This ensures robots are well-prepared for diverse real-world scenarios, significantly reducing the time required to create accurate test environments. QUARTZ

Multiple AI models help robots execute complex plans more transparently

MIT's Improbable AI Lab has developed a new AI framework, HiP (Hierarchical Planning), to aid robots in executing complex plans with more clarity and efficiency. Unlike conventional multimodal models, HiP utilizes three distinct foundation models trained on different data types: language, vision, and action. HiP's methodology is particularly effective for tasks that require multiple steps or "long-horizon goals," like household chores or construction and manufacturing processes. It breaks down these tasks into smaller, manageable sub-goals using its three foundation models: a language reasoner, a visual world model, and an action planner. This approach allows for more transparent decision-making and avoids the need for paired vision, language, and action data, which is hard to obtain. MIT

Meet The Woman Who Transformed Sam Altman Into The Avatar Of AI

OpenAI's Vice President of Global Affairs portrays the significant role in shaping the company's global outreach and regulatory approach. She has transformed Sam Altman, OpenAI's CEO, into an AI industry ambassador, organizing his meetings with world leaders and managing crisis communications. Drawing on her national security experience and time at companies like SpaceX and Facebook, she has led OpenAI's proactive engagement with policymakers worldwide. This approach differs from the reactive stance traditionally adopted by tech firms in regulatory matters. Her efforts have helped position OpenAI as a trustworthy entity in the eyes of lawmakers, influencing AI policy discussions. Despite concerns about regulatory capture, her strategy focuses on educating policymakers about AI, advocating for proactive regulation. THE WASHINGTON POST

Top 10 AI Jobs to Watch Out for in 2024

The article outlines the top 10 AI jobs anticipated to be in high demand in 2024. These roles reflect the diverse applications and rapid growth of AI across various sectors. Here's a summary:

1. Machine Learning Engineer: Focuses on designing and implementing machine learning systems and algorithms. Requires expertise in computer science, mathematics, statistics, and programming.

2. Data Scientist: Responsible for handling large datasets, employing tools like data mining and visualization. Proficiency in mathematics, statistics, and programming is key.

3. AI Research Scientist: Engages in innovative AI research, requiring a PhD or master’s degree in relevant fields and a history of published work.

4. AI Ethics Consultant: Ensures AI development aligns with ethical standards. Background in philosophy, law, sociology, or psychology, combined with AI knowledge, is necessary.

5. Robotics Engineer: Designs and maintains robots and robotic systems, often with a background in engineering fields and knowledge of robotics hardware and software.

6. AI Product Manager: Manages AI product lifecycle from conception to market, requiring business or marketing expertise and understanding of AI technologies.

7. AI Solutions Architect: Designs and implements tailor-made AI solutions, with a background in computer science or software engineering.

8. Software Engineer: Develops AI-driven software applications, necessitating a strong foundation in computer science and proficiency in programming languages.

9. Data Analyst: Extracts and analyzes data to provide insights, typically with a background in mathematics or statistics and familiarity with data analysis tools.

10. Computer Vision Engineer: Specializes in developing algorithms for processing visual information, requiring knowledge in computer science and experience with computer vision technologies.

These roles highlight the interdisciplinary nature of AI, requiring a blend of technical skills, industry knowledge, and ethical understanding.

ANALYTICS INSIGHT