AI & Startups September 23rd - September 29th

AI Product News

??? 美国艺电公司 Reveals AI-Powered Video Game Creation

Electronic Arts (EA) has unveiled a groundbreaking concept for AI-powered user-generated video game content. The reveal showcases how players can create and modify video games in real-time using natural language commands.

The Details:

?? EA's "Imagination to Creation" video demonstrates two players rapidly building complex game worlds and setting gameplay rules through voice commands without needing coding expertise.

?? The system is trained on EA's vast dataset, including 3D assets, gameplay hours, and telemetry events, enabling custom game creation models.

?? Players can remix elements from existing EA games and asset libraries to create new, unique gaming experiences.

?? Why It Matters: This could revolutionize game development by creating more interactive relationships between players and developers, opening new revenue streams, and extending game lifespans. While still in its concept stage, EA’s vision may reshape the future of video game creation.

?? 谷歌 Revolutionizes Chip Design

Google DeepMind has unveiled AlphaChip, an AI-driven system capable of designing superhuman chip layouts in just hours using reinforcement learning.

The Details:

?? AlphaChip was instrumental in designing the last three generations of Google’s Tensor Processing Units (TPUs), boosting performance and significantly reducing design cycles. ?? It uses an innovative “edge-based” graph neural network to learn relationships between chip components and generalize across multiple designs. ?? Google is sharing a pre-trained checkpoint of AlphaChip to encourage more research into AI-assisted chip design. ?? MediaTek has already adopted AlphaChip for designing advanced chips used in smartphones and other devices.

?? Why It Matters: AlphaChip sets off a self-reinforcing cycle: as AI improves chip design, those better chips train even more advanced AI models, creating a virtuous loop that will drive unprecedented acceleration in AI capabilities. The pace of chip design will become something humans can barely compete with.

?? YouTube Support Added to NotebookLM

Google’s NotebookLM just got a major upgrade, introducing support for YouTube videos and audio files, making it easier than ever to analyze multimedia content.

The Details:

?? Now, users can input public YouTube URLs and audio files, allowing them to analyze video lectures, tutorials, and more alongside text sources.

?? NotebookLM uses Gemini 1.5’s multimodal abilities to summarize key concepts and transcribe audio from videos.

?? A new sharing feature enables users to generate public links for Audio Overviews, enhancing collaboration.

?? These updates are designed to streamline tasks like creating study guides, analyzing multiple perspectives, and extracting key insights from a variety of media formats.

?? Why It Matters: With access to YouTube’s vast collection of content, NotebookLM is evolving into a powerful tool for consuming hours of video and audio in mere minutes. From educational videos to in-depth tutorials, Google’s viral hit now brings AI-powered efficiency to a whole new level.

?? Meta Unveils AR xAI Glasses, New Models, and More

At the Connect 2024 conference, Meta made waves with several major AI announcements, showcasing the future of AR, AI models, and innovative features for creators.

The Details:

?? Orion AR Glasses: Weighing under 100g, these advanced AR glasses feature voice control, hand tracking, and a wide field of view, representing over a decade of development at Meta.

?? Llama 3.2: Meta’s new vision model supports both images and text, with 11B and 90B parameter versions, designed for AI advancements across platforms.

?? Llama for On-Device: Meta also launched 1B and 3B models, optimized for smartphones and potential future AR glasses.

?? Instagram AI Updates: Automatic video dubbing, lip-syncing in any language, and AI-generated content for Feeds with the ‘Imagined for You’ feature.

?? Voice Mode: A new AI-driven voice interaction feature, similar to ChatGPT’s Voice Mode, will be integrated across Messenger, WhatsApp, Facebook, and Instagram DMs.

?? Why It Matters: Connect 2024 proves Meta’s relentless innovation. From the groundbreaking AR glasses to bringing Llama’s AI to the masses, Meta continues to push boundaries. With 500 million monthly active Meta AI users now able to access Voice Mode across platforms, the company is cementing its position at the forefront of the AI and AR revolution. Never bet against Zuck!

?? OpenAI AI's Advanced Voice Mode Launches for All Subscribers

The wait is over! This week, OpenAI is expanding access to its much-anticipated Advanced Voice Mode to all ChatGPT Plus subscribers—estimated at over 11 million users. Previously available only to a select few testers, the voice feature is now rolling out for everyone.

Why It’s Important: Advanced Voice Mode takes conversational AI to the next level. Unlike traditional chatbots, this mode simulates a real, engaging conversation where users can interject, change topics seamlessly, and even hear filler words and natural pauses. It's a more immersive experience, with human-like intonations and emotions.

What’s New? OpenAI has fine-tuned the system, introducing features such as custom instructions, memory, and five new voices with improved accents. These updates make interactions smoother and more personalized, ensuring the final product feels polished.

How to Try It: If you're a subscriber, expect a notification in your ChatGPT app by the end of the week to activate this feature!

Capital Watch: This Week's Funding Highlights

?? UJET Raises $76M Series D to Revolutionize Call Centers with AI Agents

Ujet, a call center software startup, has raised $76 million in a Series D round led by Sapphire Ventures, with participation from GV, Kleiner Perkins, and Citi Ventures. The funding values Ujet at $500 million, up from $455 million in its 2020 Series C, as investors eye the potential of generative AI to disrupt the multi-billion-dollar customer service industry.

The Details:

• Funding Amount: $76 million raised in a Series D round, bringing Ujet’s valuation to $500 million.
• Investor Support: Sapphire Ventures led the round, with continued backing from GV, Kleiner Perkins, and Citi Ventures.
• AI Integration: Ujet’s platform leverages AI to identify trends, flag customer complaints, and automate processes, helping enterprises streamline their customer service operations.
• Generative AI Potential: With CRM giant Salesforce committing $1 billion to AI innovation, Ujet is well-positioned to capitalize on the growing interest in AI-powered customer service solutions.
• Leadership Insight: Ujet’s Co-CEO Vasili Triant emphasized the need to balance AI with non-AI features, despite investor focus on generative AI. “We were highlighting the new engine we were building, but they really wanted the message and focus to be different,” Triant noted.

?? Why It Matters: Ujet’s strategic use of AI agents in the call center space is gaining traction as businesses increasingly look to automate customer service operations. With a new influx of capital and a fresh focus on generative AI, Ujet is primed to lead the charge in transforming the $40 billion call center industry.

?? Qodo (formerly Codium) Raises $40M Series A to Bring Quality-First Code Generation and Testing to Enterprises

Qodo, formerly known as CodiumAI, has raised $40 million in a Series A funding round led by Susa Ventures and Square Peg, with participation from Firestreak Ventures, ICON Continuity Fund, and seed investors TLV Partners and Vine Ventures. The startup is revolutionizing AI-powered code generation with a unique focus on quality-first code, integrating testing and issue detection throughout the software development lifecycle.

The Details:

• Funding Amount: $40 million raised in Series A funding, bringing Qodo’s total funding to $50 million.
• Enterprise Adoption: Over 1 million developers and several Fortune 100 companies have adopted Qodo’s enterprise platform since its launch 18 months ago.
• Integrated Tools: Qodo offers IDE extensions for Visual Studio Code and JetBrains, as well as a git agent for GitHub, GitLab, and BitBucket. Additional tools include a Chrome extension and a CLI tool.
• Task-Aware Coding: Developers can describe the issue they’re trying to solve in natural language, allowing Qodo’s code-completion tool to tailor its suggestions based on that context.
• Lifecycle Coverage: Qodo supports the entire software development lifecycle with tools that help reduce bugs at various stages—from code writing to GitHub commits, and even during code reviews.
• Future Expansion: Qodo is planning to launch a Qodo UX service that simulates human interactions to test a company’s website UI for bugs, expanding beyond unit and functional testing.

?? Why It Matters: Qodo’s quality-first approach to code generation and testing sets it apart in a crowded field. By focusing on integration across the development lifecycle, the platform ensures higher code quality and smoother workflows for developers at large enterprises. With its eyes set on UI testing in the near future, Qodo aims to cover all facets of the development and testing process, positioning itself as a key player in the enterprise AI development space.

?? Nurix AI raises $27.5 million in funding round led by Accel, General Catalyst

Nurix AI, a startup founded in 2024, has raised $27.5 million in Series A funding led by Accel and General Catalyst. The company is focused on building AI solutions that reinvent workflows and customer interactions for enterprises by integrating generative AI (GenAI) with human-augmented systems.

The Details:

• Funding: Nurix AI secured $27.5 million in Series A funding, with Accel and General Catalyst leading the round.
• AI Solutions: The startup specializes in custom AI agents with human-like voice and reasoning capabilities to streamline enterprise operations while maintaining personalized customer interactions.
• Founder: Mukesh Bansal , founder of Myntra and CultFit , leads Nurix AI and aims to position the company at the forefront of AI transformation.
• Expansion Plans: The funding will be used to scale technology and business operations across Asia and North America, focusing on AI research, development, and partnerships with hardware and product companies.
• First Offering: Nurix AI’s initial product will target the business process outsourcing (BPO) industry, aiming to merge AI-driven efficiency with human oversight.

?? Why It Matters: Nurix AI’s blend of AI and human collaboration seeks to redefine enterprise operations, particularly in the BPO industry. With India being a global hub for services, Nurix AI is positioned to accelerate AI adoption and innovation in customer service across the globe.

?? Prepared Secures $27M to Modernize 911 Emergency Response with AI

New York-based Prepared, an AI company transforming emergency communications, has raised $27 million in a Series B funding round led by Andreessen Horowitz, along with First Round Capital, M13, and NewView Capital. This round brings Prepared's total funding to $57 million as the company looks to modernize outdated 911 call centers across the U.S.

The Details:

• Funding Amount: $27 million raised in Series B, bringing total funding to $57 million.
• Outdated Systems: Many U.S. emergency response systems still use legacy technology built for landlines, despite the fact that 80% of 911 calls now come from wireless devices.
• AI Platform: Prepared’s AI-powered platform provides real-time data access—such as live location, photos, videos, and texts—directly to 911 call takers through an easy-to-adopt, browser-based solution.
• Scale: Prepared is currently in use across 1,000 agencies in 49 states, covering 30% of the U.S. population.
• Innovations: Recent developments include AI-generated translations that convert Spanish 911 calls to English in real-time, eliminating the need for third-party interpreters, and Apple’s Emergency SOS integration for sharing live video and texts with dispatchers.

?? Why It Matters: Emergency response technology has long lagged behind other sectors, and Prepared’s AI-driven solutions aim to bridge that gap. With this funding, Prepared is on a mission to enhance the speed and accuracy of 911 systems, empowering first responders and modernizing public safety infrastructure across the U.S.

?? Convergence Secures $12M to Develop Personal AI Agents with Long-Term Memory

London-based startup Convergence has raised $12 million in pre-seed funding to advance personal AI agents that use long-term memory for continuous learning, much like humans do. The funding round was led by Balderton Capital, with participation from Salesforce Ventures and Shopify Ventures.

The Details:

• Funding Amount: $12 million in pre-seed funding, led by Balderton Capital, with contributions from Salesforce Ventures and Shopify Ventures
• AI Capabilities: Convergence's flagship product, Proxy AI, is built for employees and consumers. It can automate repetitive tasks for employees and assist consumers with everyday activities like booking trips and ordering groceries.
• Memory Management: Proxy AI can manage its own memories, deciding in real-time what information to store or forget. This capability allows it to continuously learn and adapt during interactions.
• AI Technology: The technology is based on Large Meta Learning Models (LMLMs), which allow AI agents to "learn to learn" by extracting patterns and strategies across various tasks.
• Long-Term Memory: The startup’s unique approach incorporates memory directly into the model, teaching it not only to predict the next step but also how to manage memory by storing, remembering, or deleting information efficiently.
• Team: Founded by machine learning engineers Marvin Purtorab (CEO) and Andy Toulis (CTO)—both of whom previously worked at Shopify—the team includes alumni from Google DeepMind, Meta, and OpenAI.
• Beta Launch: Convergence’s Proxy AI agent is launching in beta today, with limited spots available.

?? Why It Matters: Convergence’s ability to create AI agents with long-term memory brings a revolutionary shift in how AI interacts with real-time information. By continuously learning and managing its memory autonomously, Proxy AI sets itself apart in an increasingly competitive AI landscape, paving the way for more intelligent, adaptable digital assistants for both consumers and enterprises.

?? Scribenote Raises $8.2M in Andreessen Horowitz-Led Funding Round

Scribenote, an AI-driven startup that automates the creation of medical records for veterinarians, has raised $8.2 million in an early-stage funding round led by Andreessen Horowitz, with additional participation from Inovia Capital and Velocity Fund. The startup, founded in 2019, is revolutionizing veterinary documentation, having automated over 1.5 million medical records in less than a year.

The Details:

• Funding Amount: $8.2 million raised in an early-stage round, led by Andreessen Horowitz.
• Additional Investors: Inovia Capital and Velocity Fund also participated in the round.
• AI Innovation: Scribenote uses artificial intelligence to automate the documentation of veterinary medical records, significantly streamlining the process for veterinarians.
• Market Potential: Startups in the AI space have attracted massive investor attention in 2024, with several companies raising over $1 billion. This underscores the growing importance of AI across industries, including healthcare and veterinary services.
• Industry Growth: Andreessen Horowitz's interest in AI continues to expand, as seen with its recent participation in a $6-billion Series B round for Elon Musk's AI startup, xAI, in May.

?? Why It Matters: Scribenote is tapping into the power of AI to make veterinary practices more efficient by automating medical documentation. With new funding from top-tier investors, the startup is poised to scale its solution, offering AI-driven tools that save veterinarians valuable time and reduce administrative burdens.

?? jhana Raises $1.6 Million Seed Funding to Develop Proprietary Legal Datasets

Bengaluru-based AI paralegal startup Jhana.ai has raised $1.6 million in seed funding. The startup, founded in 2021 at Harvard, is focused on building AI-powered legal assistants to streamline legal work for law firms, lawyers, and in-house legal teams.

The Details:

• Funding: Jhana.ai secured $1.6 million in its seed round, which began closing in December 2023. The round was led by Girish Mathrubootham (Freshworks) and Manav Garg (Eka Software), along with several other prominent investors.
• AI for Legal Work: The startup aims to develop proprietary legal datasets and AI models to assist legal professionals with tasks such as document review, legal research, and contract management, offering faster and more accurate outputs.
• Investor Participation: Notable investors include angel investor Shyamal Anadkat (Applied AI team at OpenAI), Harshil Mathur and Shashank Kumar (Razorpay founders), and Kunal Shah (CRED founder).
• Expansion Plans: Jhana.ai will use the funds to expand its dataset development and hire top-tier researchers in both law and AI. The India portion of the funding round is still being finalized and will be publicly announced soon.
• Vision: Jhana.ai’s mission is to improve legal work by integrating AI-driven solutions, enabling faster, easier, and more accurate legal processes.

?? Why It Matters: Jhana.ai’s innovative approach to AI-powered legal assistance has the potential to revolutionize the legal industry. With strong backing from influential investors and the development of proprietary datasets, the startup is set to drive efficiency and accessibility in legal services, particularly in India’s burgeoning legal tech space.

?? Youni ?? Raises €900K (~$965,000 USD) to Expand AI-Powered University Admissions Platform

Romanian edtech startup Youni, which streamlines the university admissions process through artificial intelligence, has raised €900,000 (~$965,000 USD) in a seed funding round. The round was led by Czech accelerator Soulmates Ventures with a €600,000 (~$644,000 USD) contribution, alongside Romania’s Early Game Ventures, which contributed €300,000 (~$322,000 USD).

The Details:

• Funding Amount: €900,000 (~$965,000 USD) raised in a seed round, with Soulmates Ventures contributing €600,000 (~$644,000 USD) and Early Game Ventures contributing €300,000 (~$322,000 USD).
• Platform Impact: Youni’s platform helps students worldwide navigate the complex admissions process for top universities, boasting a 100% acceptance rate for applicants. The AI system supports over 6,000 students globally and helps universities connect with qualified candidates.
• Founder’s Insight: Andrei Nicolae, Founder and CEO of Youni, noted that his own challenges with the university application process inspired the creation of the platform, which uses AI to recommend study fields and universities tailored to each student's goals and achievements.
• AI for Universities: Youni’s AI helps universities efficiently identify and recruit candidates, easing the workload of admissions teams and providing tools for better resource allocation and scholarship distribution.
• Global Reach: With partnerships in Germany and Switzerland, Youni has already secured $52 million USD in scholarships for students, helping them gain access to world-class education. The startup plans to expand into the US, UK, and Australia, with a goal to assist one million students within the next year.

?? Why It Matters: Youni is transforming the global education landscape by leveraging AI to simplify university admissions for both students and universities. With new funding, the platform is set to scale further, aiming to break down barriers to higher education, particularly for disadvantaged students, while making the admissions process more efficient for institutions.

AI IPO Boom: Unlocking the Future of Artificial Intelligence on Wall Street

?? Cerebras Systems Files for US IPO: Gearing Up to Rival Nvidia in the AI Chip Market

?? Revenue Surge: Cerebras reported a 3x revenue increase in 2023, hitting $78.7M, compared to $24.6M in 2022. The company's loss narrowed to $2.92 per share from $4.28 in the previous year.

?? Strategic Partnership: 83% of revenue came from AI tech group G42, owned by Abu Dhabi's Mubadala. This partnership is pivotal for Cerebras' growth strategy.

?? Market Projections: Cerebras predicts the AI computing market will grow from $131B in 2024 to a whopping $453B by 2027, signaling massive potential for expansion.

?? IPO Target: Cerebras is eyeing the Nasdaq under ticker CBRS, with Citigroup and Barclays leading the charge. Sources indicate they might raise up to $1B.

?? Why It Matters: With AI-related IPOs heating up, Cerebras is set to ride the wave of investor enthusiasm, aiming to challenge Nvidia's dominance in the AI chip space. Their AI hardware powers cutting-edge innovations, helping shape the future of AI computing.

?? Read More: Full SEC filing

?? Next Steps: Stay tuned for Cerebras’ public debut, as the AI race accelerates and the company positions itself for a major breakthrough!