Stable Diffusion creator Stability AI confirmed a $101M fundraise led by Coatue, Lightspeed Venture Partners, and O'Shaughnessy Ventures LLC. The oversubscribed round values the London startup at around $1B, according to Bloomberg. More: - Stability, an "independent multimodal AI company," released its popular open-source text-to-image generator, Stable Diffusion, in August.
- Since then, the open-source neural network has been downloaded by more than 200,000 developers.
- Its DreamStudio API for Stable Diffusion now claims more than 1 million registered users, who have created more than 170 million images.
- The company claims 10 million daily users "across all channels."
- The startup said it will use the fresh funding to speed up the development of open AI models for consumers and enterprises.
- This includes models for generating images, video, audio, languages, and 3D objects.
- It also plans to grow from 100 employees to 300 over the next year.
- In a release, founder and CEO Emad Mostaque emphasized the importance of making AI technology open and publicly accessible.
TECHCRUNCH 
| |
AI pair programmer GitHub Copilot has written 40% of all code for programmers during its beta testing over the last year, according to GitHub CEO Thomas Dohmke. The AI-powered code suggestion tool was made generally available to individual developers in June. Background: - Trained on large amounts of open-source code, Copilot offers suggestions based on the developer's input, including so-called "boilerplate" repetitive code.
- It uses OpenAI technology to suggest lines and functions, as well as other ways to solve problems, write tests, and find new APIs.
New functionalities: - According to Microsoft CEO Satya Nadella, the tool enables developers to code at least 50% faster.
- He said AI-assisted coding will "fundamentally change" how software is developed.
- Microsoft, which owns GitHub Copilot, is also experimenting with new functionalities to explain and translate code, Nadella added.
Not a replacement: - The 40% marks an increase from the original 35% of coding suggestions that developers originally took from Copilot, Dohmke said.
- According to Dohmke, the aim of Copilot is to help developers maintain focus and have more time for higher-value tasks than rewriting existing code.
- "Copilot today can't do 100% of the task," Dohmke said. "It's not sentient. It can't create itself without user input."
CNBC
| |
 A message from IMPACT Most marketers know that influencer marketing can really pay off for a brand. Many have waded into influencer marketing, perhaps recruiting a celebrity and a few large influencers. Yet relatively few have systematically looked at the world of influencers and asked, “What’s the best way to manage a portfolio of influencers — both large and small?” To optimize your influencer program, start off with a framework — a way to classify influencers. But what types of influencers are even out there? Use this ebook as a guide to understand: - how to engage and manage different types of influencers
- figure out which influencer is the right fit for you
- what will yield the best ROI for your program.
Download the ebook
| |
AI startups and scaleups are projected to raise $47.5B in global investments this year, down from a high of $111.4B in 2021, according to Forbes and the latest state of AI report. The annual report is produced by venture capitalists Nathan Benaich of Air Street Capital and Ian Benaich of Plural. A summary of their key takeaways is below. Smaller AI labs gain ground: - According to Benaich and Benaich, new and independent research labs are open-sourcing the AI output of bigger labs.
- The smaller labs are releasing more state-of-the-art research as compute costs drop and access becomes more available.
- Smaller labs such as Stability.ai and Midjourney are publicly releasing their own text-to-image models, showing a "fundamental shift" in the previously accepted idea that larger well-resourced labs would reign in AI research.
- Nvidia continues to lead the market in AI hardware and has enabled this acceleration to occur.
- Large AI labs now employ an estimated 300 safety researchers, a noted increase from last year's 100 researchers.
A world view: - China has a significant lead over the U.S. in AI areas concerning security and politics, such as object detection, surveillance, and autonomy.
- In 2022, global private companies using AI are projected to bring in 36% less money than in 2021 but above 2020 levels, coinciding with worldwide investments.
- The U.S. has the most "unicorn" AI startups, gaining 292 unicorns worth a combined $4.6 trillion in 2022.
- More talent from top Tier 1 labs is shifting to entrepreneurial endeavors. Some big tech firms, such as Google, are losing researchers who have launched their own AI startups.
FORBES
| |
 Ambi Robotics, a seller of AI-powered robotic arms and software for warehouses, said it has raised $32M in additional funding. The startup's parcel-sorting systems automate package processing for logistics and fulfillment operations. More: - Since its formation in 2018, Ambi has released two robotic arms. The AmbiSort automatically organizes merchandise, while the AmbiKit picks up items from boxes and conveyor belts to place them in a package.
- The company said its Ambisort A-Series V3, with suction cups and a camera vision system, can sort through 400 parcels per hour.
- Ambi also developed AmbiOS, an AI operating system for robotic arms that it says speeds up training for the robot to perform specific tasks unique to each warehouse.
- Each customer pays upfront for the robot and a monthly subscription fee for the software.
Zoom-out: - In March, Ambi and shipping company Pitney Bowes announced a $23M expansion deal to bring 80 more of the robotic arms to its parcel sorting centers.
- Pitney Bowes was also an investor in Ambi's $32M round, which was led by Tiger Global and Bow Capital with participation from Ahren.
BLOOMBERG
| |
Gaming giant Activision Blizzard has filed a patent for AI technology that automatically creates unique soundtracks for individual players during multiplayer games. The system would be able to adapt its music output based on each player's actions and activities. More: - According to the patent, the system would be capable of identifying "a mood" of a game based on event profiles.
- It would then use machine learning to modulate elements of an audio segment based on this mood, creating an original music clip tuned for that player.
- The song's intensity would also go up as a multiplayer match heats up.
- Activision noted a need for AI systems that can alter music for each player based on their skills, experience, gaming events, reactions, input, and movement during events.
- This could make games "more immersive" and "more enjoyable" with new customizable features, it noted.
GAMESPOT
| |
AI Events and Conferences: - ML:Integrity conference: Oct. 19, virtual.
- IROS 2022: International Conference on Intelligent Robots and Systems: Oct. 23-27, Kyoto, Japan.
- European Conference on Computer Vision: Oct. 23-27, Tel Aviv and online.
- The AI Summit Silicon Valley: Nov. 3-4, Santa Clara, Calif.
- MLOps Summit: Nov. 9-10, London.
- NeurIPS, Conference on Neural Information Processing Systems: Nov. 24-Dec. 4, New Orleans and virtual.
- Scaling AI Summit: Nov. 30-Dec.1, Silicon Valley.
- AI & Big Data Expo Global: Dec. 1-2, London and online.
- The AI Summit New York: Dec. 7-8, New York City.
- World Artificial Intelligence Cannes Festival: Feb. 9-11, 2023, Cannes, France, and online.
- Data Science Salon: AI & Machine Learning in the Enterprise: Feb. 21-22, 2023, Austin, Texas, and virtual.
- The European Chatbot & Conversational AI Summit: March 15-16, 2023, Edinburgh, U.K., and online.
| |
QUICK HITS - You might have an amazing product and team, but you'll struggle with your next funding round if you're weak on this one metric.*
- Neuchips, a developer of AI inference chip platforms for deep learning recommendations, raised $20M in a series B2 funding round.
- In a Capterra study, 63% of polled marketing professionals said they use AI tools for email marketing, 58% for advertising (58%), and 57% for data analysis. A lower percentage — 33% — said they use AI for content generation.
- Adobe Express for Education, its free design app for schools, has added an AI feature that suggests relevant templates based on existing content.
- AI researcher Janelle Shane prompted OpenAI's Dall-E to generate humorous images of the most popular Halloween candy in every state.
- Unlock profitable growth. Get Northbeam and discover the true impact of your marketing spend.*
*This is a sponsored post.
| |
UPCOMING EVENTS - October 14 - Learn how leaders from Slack, Shopify, and Stripe are improving developer productivity, workflow, and collaboration in their engineering orgs. (Register Here) *
- October 19 - AMA with Zecca Lehn (Responsibly VC) (Register Here)
- October 25 - Meet Our Fund 4, an Inside.com Summit (Register Here)
- November 02 - AMA with Bill Glenn (Executive Chairman at Crenshaw Associates) (Register Here)
- November 09 - Live AMA with Yashar Behzadi (CEO and Founder of Synthesis AI)* (Register Here)
- December 07 - Live AMA with Deividi Silva ( Head of Developer Relations at Gun.io) (Register Here)
*This is a sponsored listing.
| |
 |  | Beth is a writer and editor at Inside.com covering artificial intelligence and daily technology news. You can reach her at Beth.Duckett@inside.com. |   | | Editor | Aaron Crutchfield is based in the high desert of California. Over the last two decades, he has spent time writing and editing at various local newspapers and defense contractors in California. When he's not working, he can often be found looking at the latest memes with his kids or working on his 1962 and 1972 Fords. | |
| |
Comments
Post a Comment