The Gist

  • An AI Assistant. Microsoft unveils Copilot.
  • Finding the right words. Google USM understands 2,000 languages.
  • AI funding news. SOCi and Humane raise millions

It’s been an exciting week in AI as OpenAI released developer APIs for ChatGPT and Whisper AI models, giving developers the ability integrate OpenAI's advanced tech into their own applications. And a battle began brewing in the smart CRM arena with HubSpot announcing the release of and the very next day, Salesforce introduced Einstein GPT, both powered by ChatGPT and each a new generative AI CRM technology set to compete in the smart CRM space.

In other AI news...

Microsoft Introduces AI Assistant

On March 6, Microsoft unveiled a series of AI product updates across its business applications portfolio, including the launch of the new Microsoft Dynamics 365 Copilot — providing interactive, AI-powered assistance across business functions.

According to company officials, Dynamics 365 Copilot is an AI-powered assistant that provides personalized support to customers, helping businesses to enhance their customer engagement and satisfaction, using natural language processing and machine learning capabilities to understand customers' needs and provide appropriate solutions. The tool can be integrated into various communication channels, including websites, mobile apps and messaging platforms, providing customers with support across multiple touchpoints. Copilot can also assist businesses in identifying patterns in customer inquiries, enabling them to proactively address potential issues. Microsoft plans to roll out Dynamics 365 Copilot to more businesses in the coming months.

Google Universal Speech Model (USM) for AI Language

Google researchers have developed a new universal speech model (USM) that has achieved state-of-the-art performance on multiple speech recognition benchmarks. The model is capable of understanding and transcribing speech in more than 2,000 languages and can be trained using data from any language. The USM can also adapt to different accents and dialects, making it a versatile tool for speech recognition applications.

According to company officials, the USM uses a self-supervised learning approach, which allows the model to learn from unlabeled data, making it more efficient than traditional supervised learning models that require a large amount of labeled data. The USM is also highly accurate, achieving an average word error rate of 1.4% on a benchmark dataset.

The USM has many potential applications, including speech-to-text transcription, voice-controlled devices and language translation. Google has already used the USM to develop a new speech recognition system for its Android operating system, which has shown significant improvements in accuracy. The company plans to make the USM available to developers through its machine learning platform, TensorFlow, and hopes that it will help to drive advances in speech recognition technology.

SOCi Raises $120M for AI-Powered Marketing

SOCi Inc., a marketing platform for multilocation brands, has announced it raised a total of $120 million led by JMI Equity (JMI), a growth equity firm. According to company officials, the funds will go toward investing in a line of AI marketing products called "Genius" that will leverage advanced artificial intelligence and machine learning model. Through an OpenAI integration into the company’s review response management tool, ChatGPT’s natural language model will enable instant, intelligent responses to online reviews.

In addition to its investment in AI technology, SOCi will use the funding to expand into new markets, including manufactured goods.

"SOCi’s customers have thousands of marketing decisions to make across hundreds or thousands of locations. The opportunity for optimization is incredible, but at such a scale it’s stunted by true human capacity issues," Afif Khoury, CEO and co-founder of SOCi, said in a statement. "Our Genius products are the answer, as they will leverage sophisticated data science models to drive recommendations and automate data-driven decisions across the organization."

Related Article: ChatGPT Is Back, Snapchat Welcomes Chatbot, More AI News

Digital Wave Technology Acquires AI Software Company, GoalProfit

Digital Wave Technology Inc., an omnichannel product platform, has announced its acquisition of GoalProfit Inc., an Al and optimization solutions company based in Austin, Texas, and Helsinki, Finland. Originally forming a partnership in early 2022, Digital Wave Technology and GoalProfit focused on the development and execution of Digital Wave’s omnichannel platform utilizing artificial intelligence including generative AI.

 “We’ve made significant progress in a relatively short period of time, aided by our domain-experienced team,” Lori Schafer, CEO of Digital Wave Technology, said in a statement. “At the same time, we’re acutely aware that positioning for the next major step in growth requires world-class acumen in data science and low-code, no-code analytical architecture.”

Pega Platform Integrates Generative AI Technology

Pegasystems Inc., a low-code platform provider, has announced plans to extend its use of AI in its platform with a set of new generative AI capabilities across Pega Infinity. Powered by generative AI models like those behind OpenAI's ChatGPT, Pega is designing these features to enable users to use generative AI with enterprise governance to complete tasks using natural language prompts.

According to company officials, the new capabilities will be available with the launch of Pega Infinity '23 in Q3 2023, and the new generative AI capabilities will be showcased at PegaWorld iNspire, the company's annual conference June 11-13 in Las Vegas.

“The world has been captivated by the endless possibilities of generative AI, and now Pega is working on unlocking generative AI's potential for widespread use in the enterprise with the upcoming launch of Pega Infinity '23," Kerim Akgonul, chief product officer at Pega, said in a statement. "These capabilities will enhance our existing AI and automation capabilities in new ways we have only begun to explore. By combining generative AI and Pega, our clients will be able to dramatically accelerate their path to digital transformation.”

Learning Opportunities

Brain Waves Create AI Images

Scientists have developed a new artificial intelligence (AI) technique that can generate highly accurate images based on brain activity. The technique, called stable diffusion (StableD), analyzes brain waves recorded by an electroencephalogram (EEG) and generates images that correspond to the subject's visual perception. The images are created using a neural network that has been trained on a large dataset of images and can predict how different patterns of brain activity correspond to different visual stimuli. According to researchers, the StableD technique has shown promising results in a study of 10 subjects, with an average accuracy of 93%.

The ability to translate brain waves into images could have significant implications for medical research and neuroscience. The technique could be used to study how the brain processes visual information and to better understand the neural mechanisms underlying perception. It could also be used to develop new diagnostic tools for neurological disorders, such as stroke or traumatic brain injury, that affect a patient's visual perception. Additionally, the technique could have applications in the field of brain-computer interfaces, enabling people with paralysis or other disabilities to communicate using their brain activity.

The StableD technique represents a significant advance in the field of AI-generated images and has potential applications beyond neuroscience. The technique could be used to create realistic simulations for video games or virtual reality environments or to generate images based on other types of biological data, such as gene expression or protein activity. However, researchers caution that the technique is still in the early stages of development and requires further validation and refinement before it can be used in practical applications.

Related Article: OASIS Makes Content out of Jumbled Thoughts, Google Bard's $100B Blunder, More AI News

Omneky Launches Creative Assistant Utilizing ChatGPT API

Omneky, a generative AI advertising platform, has announced the launch of AI-infused Creative Assistant, a chat-based interface using OpenAI's new GPT-3.5 Turbo. The Creative Assistant utilizes user input to develop a creative brief, complete with AI generated visual storyboard and mood board.

Omneky also recently launched a different new feature, the Custom AI-Generated Product Imagery tool, which has two functions — producing rapid iterations of product images based on catalog photos and generating custom image backgrounds. In addition to Omneky's proprietary algorithms, this tool also utilizes Stable Diffusions' Dreambooth, Dall-E's inpainting capability and other deep learning models.

"The launch of Omneky's Creative Assistant is the next step in reimagining human computer interface with AI,” Hikari Senju, founder and CEO at Omneky, said in a statement. "We believe that data-driven, AI-powered chat is the next modality for creating advertising content. The barriers to creating content are still too high for most people, requiring the learning of complex image and video editing tools and techniques. AI has the potential to reimagine the content creation process so anyone, no matter their skill level, can create beautiful ads through a conversational interface. Similar to how someone would have a conversation with an artist to commission a work of art, the future of advertising content will be created via a conversation with Omneky's data-driven AI.”

Humane raises $100M

Humane Inc., an AI software startup, has announced a $100 million Series C funding round led by Kindred Ventures. Founded in 2018 by ex-Apple team Imran Chaudhri and Bethany Bongiorno, Humane will use the funding to continue work on its software platform and consumer device built from the ground up for artificial intelligence (AI).

The round included participation from existing investors, Tiger Global, Valia Ventures, Qualcomm Ventures, Forerunner Ventures, Lachy Groom and Sam Altman, founder of OpenAI — as well as participation from new investors and partners including Hico Capital (US based investment arm of SK networks), Microsoft, Volvo Cars Tech Fund, LG Technology Ventures, Top Tier Capital, Hudson Bay Capital and Socium Ventures.

 “Our first device will enable people to bring AI with them everywhere. It’s an exciting time, and we’ve been focused on how to build the platform and device that can fully harness the true power and potential of this technology,” Imran Chaudhri, co-founder at Humane, said in a a statement. “We are extremely proud to announce this industry leading lineup of partners and collaborators, and we are thrilled to work closely with them to bring Humane’s vision to market. We are at the beginning of the next era of compute and believe that together we can begin the journey to fundamentally reshape the role of technology in people’s lives.”

AI Tweet of the Week

With the recent release of APIs for OpenAI's ChatGPT and Whisper AI, developers can now integrate the models into their apps and products — giving them access to some of the most cutting-edge machine learning tools available. Rowan Cheung, creator of the AI newsletter @therundownai, recently tweeted that more than 100 new products have debuted since the the ChatGPT API release — and shared 20 of the trending tools. 

fa-regular fa-lightbulb Have a tip to share with our editorial team? Drop us a line: