Categories
AI

Meta announces plans to build an AI-powered ‘universal speech translator’

Meta, the owner of Facebook, Instagram, and WhatsApp, has announced an ambitious new AI research project to create translation software that works for “everyone in the world.” The project was announced as part of an event focusing on the broad range of benefits Meta believes AI can offer the company’s metaverse plans.

“The ability to communicate with anyone in any language — that’s a superpower people have dreamed of forever, and AI is going to deliver that within our lifetimes,” said Meta CEO Mark Zuckerberg in an online presentation.

The company says that although commonly spoken languages like English, Mandarin, and Spanish are well catered to by current translation tools, roughly 20 percent of the world’s population do not speak languages covered by these systems. Often, these under-served languages do not have easily accessible corpuses of written text that are needed to train AI systems or sometimes have no standardized writing system at all.

Meta says it wants to overcome these challenges by deploying new machine learning techniques in two specific areas. The first focus, dubbed No Language Left Behind, will concentrate on building AI models that can learn to translate language using fewer training examples. The second, Universal Speech Translator, will aim to build systems that directly translate speech in real-time from one language to another without the need for a written component to serve as an intermediary (a common technique for many translation apps).

In a blog post announcing the news, Meta researchers did not offer a timeframe for completing these projects or even a roadmap for major milestones in reaching their goal. Instead, the company stressed the utopian possibilities of universal language translation.

“Eliminating language barriers would be profound, making it possible for billions of people to access information online in their native or preferred language,” they write. “Advances in [machine translation] won’t just help those people who don’t speak one of the languages that dominates the internet today; they’ll also fundamentally change the way people in the world connect and share ideas.”

Crucially, Meta also envisions that such technology would hugely benefit its globe-spanning products — furthering their reach and turning them into essential communication tools for millions. The blog post notes that universal translation software would be a killer app for future wearable devices like AR glasses (which Meta is building) and would also break down boundaries in “immersive” VR and AR reality spaces (which Meta is also building). In other words, though developing universal translation tools may have humanitarian benefits, it also makes good business sense for a company like Meta.

It’s certainly true that advances in machine learning in recent years have hugely improved the speed and accuracy of machine translation. A number of big tech companies, from Google to Apple, now offer users free AI translation tools, used for work and tourism, and undoubtedly provide incalculable benefits around the world. But the underlying technology has its problems, too, with critics noting that machine translation misses nuances critical for human speakers, injects gendered bias into its outputs, and is capable of throwing up those weird, unexpected errors only a computer can. Some speakers of uncommon languages also say they fear losing hold of their speech and culture if the ability to translate their words is controlled solely by big tech.

Considering such errors is critical when massive platforms like Facebook and Instagram apply such translations automatically. Consider, for example, a case from 2017 when a Palestinian man was arrested by Israeli police after Facebook’s machine translation software mistranslated a post he shared. The man wrote “good morning” in Arabic, but Facebook translated this as “hurt them” in English and “attack them” in Hebrew.

And while Meta has long aspired to global access, the company’s own products remain biased towards countries that provide the bulk of its revenue. Internal documents published as part of the Facebook Papers revealed how the company struggles to moderate hate speech and abuse in languages other than English. These blind spots can have incredibly deadly consequences, as when the company failed to tackle misinformation and hate speech in Myanmar prior to the Rohingya genocide. And similar cases involving questionable translations occupy Facebook’s Oversight Board to this day.

So while a universal translator is an incredible aspiration, Meta will need to prove not only that its technology is equal to the task but that, as a company, it can apply its research fairly.

Repost: Original Source and Author Link

Categories
AI

Automation is not enough: Buildings need AI-powered smarts

We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 – 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!


Buildings have been one of the most voracious users of IoT devices. Smart buildings, in particular, use connected devices to measure everything from temperature, lighting, air quality, noise, vibration, occupancy levels and energy consumption — and that’s just the very tip of the iceberg.

Building automation is big and getting bigger, with well over 6 million commercial buildings in the U.S. alone and an estimated 2.2 billion connected devices deployed. The global market for building automation systems in 2022 will reach about $80 billion. 

This type of automation relies on fleets of IoT devices. Many condition-action responses are automated; if a fire is detected, alarms are automatically triggered, often with voice instructions and fire departments are notified. That was true before the IoT; now fire alarms are connected by the Internet and secondarily via cellular communication. 

The value of IoT, in building automation specifically, is realized in two main areas:

  • The data generated by in-building devices and how it is analyzed and leveraged. 
  • The actions and management performed by building automation systems

Rich, ongoing data streams provide valuable insights into building operations, but there’s an issue: large device fleets create large volumes of data that humans alone cannot properly parse and understand. To realize the potential payoff from deploying these sensors (and cameras), artificial intelligence (AI) and machine learning (ML) is needed to continuously monitor and assess the data streams. 

Automation can’t do the job alone

Until 2020, the emphasis of smart buildings systems, including building automation, was the responsibility of facilities’ management. Then, the focus shifted to employee health and ESG initiatives, in addition to facilities management. This opened up demand for capabilities that ML enables.

An AI system can observe air quality and find correlations with occupancy limits, for instance. It can also learn how to reassign conference rooms and cubicles, relating to occupancy and ventilation, with the goals of maximizing the physical distance between employees and improving air quality, to reduce the chance of employee illness. 

AI can also help analyze the usage of water supply pipes and water temperature to warn when there is an elevated risk of legionella and other harmful pathogens. Legionella thrives in specific temperature ranges of warm water. 

The relevance of new AI-enabled capabilities does not rule out the traditional functions such as tracking and managing energy consumption. With an AI-driven platform, a building can power down areas that are not in use and try different window shade settings at different times, to minimize energy usage. Experiment and learn as it goes. This is a bottom-line issue and will become more important in 2022 due to energy prices. 

AI can even play a role in cleaning efficiency, identifying which desks have been used and which toilets have seen increased usage. In the age of COVID-19, facilities managers are focused on cleanliness. 
AI can greatly enhance systems that support physical security, too. Once a system learns what constitutes normal access and movement behavior, it can identify anomalous behavior and alert security. Other AI-driven applications can detect duress situations, abandoned objects,  recognize weapons, pinpoint shots fired—and carry out emergency lockdowns.

An intelligent infectious disease control system can learn to leverage data on local infection rates. AI systems can do things people cannot, like staring at a wall for 20 years and looking for signs of change in the concrete that could herald a pending structural collapse.

Applying AI for smart buildings

The standard starting point for a new AI-driven system is, of course, teaching it. That process begins with a foundation of data that represents the realities that the system will confront. Many will find, however, that good base training data for smart-building systems does not exist. The answer can be to create the training data by running ‘experiments’ in the physical building. 

In energy consumption, for example, you can train a system by experimentally adjusting window shades and AC based on the time of day and office occupancy, to lower AC bills without triggering a manual override. Such a system could rely on temperature sensors and occupancy readings, as well as sunlight detection. 

There are basic best practices to follow. Be scientific and rigorous when collecting ground truth datasets and collect data from multiple sources to increase confidence that your samples are representative. 

AI-driven systems can learn from the occupancy patterns of specific office areas and help reduce human error in space planning. Upgrading space is costly and preserving flexibility is vital. Space utilization and occupancy obviously became a health issue during the pandemic. Employees may now prefer to gather for conversation and coffee on an open-air balcony or patio, not in a small break room. 

Where AI-driven building management is heading

AI-powered systems can recommend changes to facilities management and allow building management to be more predictive. When it comes to reactivity, they enable a more effective response to surprise challenges as well. A recent example; before 2020, identifying employees who are running hot (fever) and reducing the probability of infection probability was not a thing, but it is within current capabilities to address this problem.

It takes careful thought and putting in the time, to get the ground truth right. Many commercial buildings have a digital twin; a virtual replica delivered by the architect to the building owner or manager. The digital twin, as a starting point, may well be a testing ground for AI-driven facilities management and smart building management.

We expect that IT, facilities management, HR and security will become more integrated and make increased use of AI. There is a range of likely benefits from joining their information silos to create data streams for AI applications. 

The importance of healthy workplaces, physical security and energy conservation makes it urgent to go beyond simple automation and develop reliable AI-based building operating systems that are founded on robust, up-to-date data. Any of these applications support a strong business case; taken together, they make a persuasive argument that facilities’ management should look at AI-driven applications for operating smart buildings and making buildings smarter. 

William Cowell de Gruchy is the founder and CEO of Infogrid.

DataDecisionMakers

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even consider contributing an article of your own!

Read More From DataDecisionMakers

Repost: Original Source and Author Link

Categories
AI

Meta launches PyTorch Live to build AI-powered mobile experiences

Join gaming leaders, alongside GamesBeat and Facebook Gaming, for their 2nd Annual GamesBeat & Facebook Gaming Summit | GamesBeat: Into the Metaverse 2 this upcoming January 25-27, 2022. Learn more about the event. 


During its PyTorch Developer Day conference, Meta (formerly Facebook) announced PyTorch Live, a set of tools designed to make AI-powered experiences for mobile devices easier. PyTorch Live offers a single programming language — JavaScript — to build apps for Android and iOS, as well as a process for preparing custom machine learning models to be used by the broader PyTorch community.

“PyTorch’s mission is to accelerate the path from research prototyping to production deployment. With the growing mobile machine learning ecosystem, this has never been more important than before,” a spokesperson told VentureBeat via email. “With the aim of helping reduce the friction for mobile developers to create novel machine learning-based solutions, we introduce PyTorch Live: a tool to build, test, and (in the future) share on-device AI demos built on PyTorch.”

PyTorch Live

PyTorch, which Meta publicly released in January 2017, is an open source machine learning library based on Torch, a scientific computing framework and script language that is in turn based on the Lua programming language. While TensorFlow has been around slightly longer (since November 2015), PyTorch continues to see a rapid uptake in the data science and developer community. It claimed one of the top spots for fast-growing open source projects last year, according to GitHub’s 2018 Octoverse report, and Meta recently revealed that in 2019 the number of contributors on the platform grew more than 50% year-over-year to nearly 1,200.

PyTorch Live builds on PyTorch Mobile, a runtime that allows developers to go from training a model to deploying it while staying within the PyTorch ecosystem, and the React Native library for creating visual user interfaces. PyTorch Mobile powers the on-device inference for PyTorch Live.

PyTorch Mobile launched in October 2019, following the earlier release of Caffe2go, a mobile CPU- and GPU-optimized version of Meta’s Caffe2 machine learning framework. PyTorch Mobile can launch with its own runtime and was created with the assumption that anything a developer wants to do on a mobile or edge device, the developer might also want to do on a server.

“For example, if you want to showcase a mobile app model that runs on Android and iOS, it would have taken days to configure the project and build the user interface. With PyTorch Live, it cuts the cost in half, and you don’t need to have Android and iOS developer experience,” Meta AI software engineer Roman Radle said in a prerecorded video shared with VentureBeat ahead of today’s announcement.

Built-in tools

PyTorch Live ships with a command-line interface (CLI) and a data processing API. The CLI enables developers to set up a mobile development environment and bootstrap mobile app projects. As for the data processing API, it prepares and integrates custom models to be used with the PyTorch Live API, which can then be built into mobile AI-powered apps for Android and iOS.

In the future, Meta plans to enable the community to discover and share PyTorch models and demos through PyTorch Live, as well as provide a more customizable data processing API and support machine learning domains that work with audio and video data.

PyTorch Live

“This is our initial approach of making it easier for [developers] to build mobile apps and showcase machine learning models to the community,” Radle continued. “It’s also an opportunity to take this a step further by building a thriving community [of] researchers and mobile developers [who] share and utilize pilots mobile models and engage in conversations with each other.”

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

AI-powered voice transcription startup Verbit secures $250M

Hear from CIOs, CTOs, and other C-level and senior execs on data and AI strategies at the Future of Work Summit this January 12, 2022. Learn more


Verbit, a startup developing an AI-powered transcription platform, today announced that it secured $250 million, bringing its total capital raised to $550 million. The round — a series E, made up of a $150 million primary investment and $100 million in secondary transactions — was led by Third Point Ventures with participation from Sapphire Ventures, More Capital, Disruptive AI, Vertex Growth, 40North, Samsung Next, and TCP.

With the fresh capital, Verbit, which is now valued at $2 billion, plans to expand its workforce while supporting product research and development as well as customer acquisition efforts. Beyond this, CEO Tom Livne said that Verbit will pursue further mergers and acquisitions and “provide enhanced value” to its media, education, corporate, legal, and government clients.

During the pandemic, enterprises ramped up their adoption of voice technologies, including transcription, as remote videoconferencing became the norm. In a survey from Speechmatics, a little over two-thirds of companies said that they now have a voice technology strategy. While they cited accuracy and privacy as concerns, 60% without a strategy said that they’d consider one within five years — potentially driving the speech and voice recognition market to $22 billion in value by 2022.

Livne cofounded New York-based Verbit with Eric Shellef and Kobi Ben Tzvi in 2017. Shellef previously led speech recognition at Intel’s wearables group, while Tzvi cofounded and served as CTO at facial recognition startup Foresight Solutions. As for Livne, who’s also a member of Verbit’s board, he was an early investor in counter-drone platform Convexum, which was acquired by NSO Group in 2020 for $60 million.

AI-powered transcription

Verbit’s voice transcription and captioning services aren’t novel — well-established players like Nuance, Cisco, Otter, Voicera, Microsoft, Amazon, and Google have offered rival products for years, including enterprise-focused platforms like Microsoft 365. But Verbit’s adaptive speech recognition tech can generate transcriptions that it claims achieve higher accuracy than its rivals.

Verbit users upload audio or video to a dashboard for AI-powered processing. Then, a team edits and reviews the material — taking into account customer-supplied notes and guidelines.

Finished transcriptions from Verbit are available for export to services like Blackboard, Vimeo, YouTube, Canvas, and Brightcove. A web frontend shows the progress of jobs and lets users edit and share files or define the access permissions for each, plus add inline comments, requesting reviews, or viewing usage reports.

“Verbit’s in-house AI technology detects domain-specific terms, filters out background noise and echoes, and transcribes speakers regardless of accent to generate … transcripts and captions from both live and recorded video and audio. Acoustic, linguistic, and contextual data is … checked by our transcribers, who [incorporate] customer-supplied notes, guidelines, specific industry terms, and requirements,” Livne told VentureBeat via email. “By indexing video content for web searches, Verbit [can help] companies improve SEO and increase their site traffic. [In addition, the platform can] provide audio visual translation to help global businesses with translations and to reach international audiences with their products and offerings.”

The transcriber experience

Like its competition, Verbit relies on an army of crowdworkers to transcribe files. The company’s roughly 35,000 freelancers and 600 professional captioners are paid in one of two ways, per audio minute or word. While Verbit doesn’t post rates on its website, a source pegs transcription pay at $0.30 per audio minute. Two years ago, transcription service Rev faced a massive backlash when it slashed minimum rates for its transcribers from $0.45 to $0.30 per word transcribed.

In some cases, pay can dip below $0.30 on Verbit, according to employee reviews on Indeed. The company reportedly started paying as low as around $0.24 cents per audio minute last year for a standard job.

Transcription platforms also don’t always have the technology in place to prevent crowdworkers from seeing disturbing content. In a piece by The Verge, crowdworkers on Rev said that they were exposed to graphic or troubling material on multiple occasions with no warning, including violent police recordings, descriptions of child abuse, and graphic medical videos.

A spokesperson told VentureBeat via email: “Currently, we employ a mix of full-time transcribers and captioners, as well as freelancers that are paid per audio minute. We’ve established a ranking system based on efficiency and accuracy to incentivize and reward freelancers with higher compensation in exchange for consistently delivering high-quality transcripts … The company’s transcribers have a support system — chat and forum — that constantly relays feedback to Verbit management, and it has a bonus program to ensure proper compensation for its top performers.”

The spokesperson continued: “In addition to competitive pay and opportunities for advancement, our staff of full-time transcribers and captioners are eligible to receive healthcare benefits … Our transcriber community follows a ranking system based on tenure and number of hours worked, allowing freelancers to earn promotions to roles such as editor, reviewer, and supervisor.”

On the subject of graphic content, the spokesperson said: “Verbit does not take on any business associated with violent or graphic content. For example, an adult entertainment company recently requested our services, but we chose not to accept them as a customer.”

Growth year

Verbit’s platform has wooed a healthy base of over 2,000 customers, bolstered by its acquisition of captioning provider VITAC earlier this year. In recent months, Verbit has pursued contracts with educational institutions like Harvard and Stanford, which have stricter accommodation standards than organizations in other sectors.

Auto captioning technologies on YouTube, Microsoft Teams, Google Meet, and like platforms aren’t beholden to the accommodations standards outlined in the Americans with Disabilities Act. In contrast, captioning must satisfy certain accuracy criteria in order to meet federal guidelines. A recent survey conducted by Verbit found that only 14% of schools provided captions as a default, while about 10% said that they only caption lessons when a student requests it.

Verbit also says that it’ll continue to explore verticals in the insurance, financial, media, and medical industries. The company — which currently has 470 employees, a number that it expects will grow to 750 by 2023 — recently launched a human-in-the-loop transcription service for media outlets and inked an agreement with the nonprofit Speech to Text Institute to invest in court reporting and legal transcription.

“With six times year-over-year revenue growth and close to $100 million in annual recurring revenue, Verbit continues to expand into new verticals at a hyper-growth pace. The shift to remote work and accelerated digitization amid the pandemic has been a major catalyst … and has further driven Verbit’s rapid growth,” Livne added. “In today’s digital era where audio and video content is a given, and many times the main method of conveying information, these AI tools are crucial to ensure that individuals and organizations of all sizes and forms can engage with their audiences and stakeholders more efficiently and effectively.”

Livne previously said that Verbit plans to file for an initial public offering in 2022.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

Grammarly raises $200M to expand its AI-powered writing suggestions platform

Join gaming leaders, alongside GamesBeat and Facebook Gaming, for their 2nd Annual GamesBeat & Facebook Gaming Summit | GamesBeat: Into the Metaverse 2 this upcoming January 25-27, 2022. Learn more about the event. 


San Francisco, California-based Grammarly, which develops an AI-powered writing assistant, today announced that it raised $200 million, valuing the company at $13 billion post-money. Baillie Gifford led the round with participation from funds and accounts managed by BlackRock. Grammarly global head of product Rahul Roy-Chowdhury says the funding will be put toward further developing the company’s technology and “accelerat[ing] efforts to help people communicate … in our digital-first world.”

As enterprises increasingly embrace digitization, the over $1.2 billion AI writing assistant market is expected to grow at a compound annual growth rate of 27.6% from 2018 to 2028. According to a survey from John Snow Labs and Gradient Flow, 60% of tech leaders indicated that their budgets for natural language processing — which encompasses technologies like Grammarly’s — grew by at least 10% compared to 2020, while a third said that their spending climbed by more than 30%.

“From changes in our tone to workplace communication shifts at large, the global pandemic has left a lasting impact on how we communicate. The shift to remote-first everything — including school and work — has only increased the need for digital communication assistance like ours,” Roy-Chowdhury told VentureBeat via email. “Poorly-written communication was already estimated to cost businesses billions annually in lost productivity — and that was well before the pandemic forced people to start communicating remotely across all kinds of new connectivity and productivity tools.”

A brief history

The brainchild of Ukrainian developers Alex Shevchenko, Max Lytvyn, and Dmytro Lider, Grammarly was founded in 2009 as Sentenceworks shortly after the sale of Lytvyn and Shevchenko’s plagiarism-checking startup MyDropBox to Blackboard. One of Grammarly’s first products was a barebones WYSIWYG editor into which users could paste text, which evolved into an app that provided spelling and syntax suggestions.

In 2012, Grammarly pivoted from a largely academic customer base to the broader consumer market, rolling out subscription plans and launching ads on Facebook, Twitter, and YouTube. In 2014, Grammarly debuted plugins for Microsoft Office that integrated its grammar checker with Outlook and Word. And in 2015, the company launched the Grammarly browser extension for Chrome and Safari. A Firefox extension followed shortly after, as did an add-on for Google Docs; the message- and email-scanning Grammarly Keyboard for iOS and Android; and a desktop client for MacOS and Windows.

Today, Grammarly offers its entry-level, English-only Grammarly Premium service for $30 a month (or $144 a year), while the enterprise Grammarly Business tier starts at $25 per user per month. Both add style and vocabulary recommendations across all platforms as well as full-sentence rewrites and a tone detector that can identify “tempers” (e.g., “aggressive,” “annoyed,” and “excited”) in emails, documents, and blogs.

“Our business model is a freemium model, in which we offer a free version of our product as well as Grammarly Premium and Grammarly Business, which are paid upgrades,” a Grammarly spokesperson told TechCrunch in a 2019 interview. “The only way Grammarly makes money is through its subscriptions … We don’t sell or rent user data to third parties for any reason, including for them to deliver their ads. Period.”

As of October, eligible Samsung smartphones running the Samsung Keyboard benefit from Grammarly’s spell-checking technology. Grammarly’s writing suggestions are built into the Samsung Keyboard, although access to the company’s premium features requires a subscription.

“Looking toward the future, expanding our critical communication support starts with launching our latest product offering, Grammarly for Windows and Mac. This is a new way to experience Grammarly across any desktop or online application, and a big step in bringing our real-time writing assistance to more places people write — like Microsoft PowerPoint and Slack’s native app, as just a few examples,” Roy-Chowdhury said. “We’ll also double down on partnerships, our next frontier, to expand the reach of our writing assistance … And as more business leaders recognize the direct link between effective communication and business performance, we’re ready to meet the growing demand for Grammarly Business.”

Expanding segments

Spurred by digital transformations that accelerated during the pandemic, a larger share of companies are expected to adopt AI technologies that automatically recommend and tailor copy, particularly in marketing. According to a survey by Phrasee, 63% of marketing teams would consider investing in AI to generate and optimize ad copy. Sixty-five percent trust that AI can generate desirable brand language, the survey found, while 82% believe that their organization would benefit from data that provides insights into how consumers respond to that language.

In an effort to broaden its appeal within markets with specific editing needs, like sales, Grammarly recently introduced Grammarly for Developers, a development kit that enables companies to add Grammarly-powered recommendations to existing web-based apps. Grammarly for Developers, currently in closed beta, handles communication between apps and Grammarly’s cloud services, rendering underlines and suggestion cards, applying text transformations, and providing personal dictionaries.

“With Grammarly for Developers, we’re helping developers deliver a better writing experience to their users while saving them valuable time and resources needed to develop their own technologies,” Roy-Chowdhury added. “Our first Grammarly for Developers product, the Text Editor SDK, brings Grammarly’s communication assistance to any web-based application … Grammarly is [now] available on more than 500,000 applications and sites.”

In June, Grammarly rolled out additional features focused on enterprise communications, including support for up to 50 style guides, snippets, brand tones, and an analytics dashboard. Snippets are preset response templates for quick communication. Brand tones are “tone profiles” that specify which tones team members should use — and avoid. As for the analytics dashboard, it shows metrics that make it ostensibly easier for leaders to identify communication trends and strengths.

Grammarly has competition in Writer, Ginger, ProWritingAid, and Slick Write, which similarly provide an AI-powered writing assistant for a range of use cases. Pitted against Ginger, an English professor writing for Fast Company found that Grammarly’s tips “lack an understanding of nuance,” for example overzealously policing redundancies.

But Grammarly — which has over 600 team members, and is profitable — says it’s managed to attract 30 million users and 30,000 teams to date. New and existing customers include Frost & Sullivan, HackerOne, Lucid, Zoom, Cisco, and Expedia.

The company’s total capital stands at more than $400 million with the latest funding round.

“In 2020 alone, we delivered 1.2 trillion writing suggestions to our global user base.  Sixty-eight percent of the Forbes Global 2000 companies have at least one Grammarly user — and we expect that to increase significantly in the years to come,” Roy-Chowdhury said. “The number of Grammarly Business customers with large-dollar contracts increased by more than 250% this past year alone. We also support over 6,800 nonprofits and nongovernmental organizations from over 150 countries with a free offering, while Grammarly’s education division serves over 2,500 educational institutions with capabilities tailored to their needs.”

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

AI-powered writing assistant Writer nabs $21M

Writer, which bills itself as an AI writing assistant for marketing teams, today announced that it raised $21 million. (Writer’s total now stands at $25 million.) Insight Partners led the series A round with participation from Gradient Ventures, which Writer CEO May Habib says will be put toward customer acquisition initiatives and headcount growth.

Writer’s growth comes as marketers increasingly look to AI to bolster their customer outreach. According to McKinsey, 80% of high-performing companies have adopted AI in marketing and sales for tasks like pricing, prediction of likelihood to buy, and customer service analytics.

“Our vision is great writing for everyone,” Habib said in a statement. “Most teams don’t have the editorial resources to ensure strong writing and consistent messaging across large amounts of content, so we provide a seamless way to help everyone at a company write well, write fast, and be on-brand.”

AI-powered suggestions

San Francisco, California-based Writer, formerly Qordoba, was founded by Habib and Waseem Alshikh in 2015. Alshikh was previously the CEO at iMena, a holding company with news, ecommerce, and classified ads businesses operating in the Middle East and North Africa. Habib was VP at one of the world’s largest sovereign wealth funds, where she was the first employee on the technology investment team.

Writer

The two started Writer out of a mutual desire to build software that helps companies write more clear, consistent marketing copy. Leveraging AI, Writer’s platform delivers guidelines that help organizations align content spanning communications, marketing, product, and human resources documents.

On the AI side, Writer employs an engine that evaluates things like plagiarism, sentence complexity, tone, paragraph length, spelling and grammar, formality, active voice usage, and other key metrics. Beyond this, the platform lets companies create a “single source of truth” for brand terms that users can build, edit, and share. For example, teams can provide examples of usage with descriptions and guidance and use tags, filters, statuses, and edit history to organize terms into a taxonomy across apps including Chrome, Microsoft Word, and Figma.

“From single sentences to page-long templates, make it easy for your team to reuse your approved content,” Writer explains on its website. “Preserve formatting, lists, and links, and include variable placeholders for dynamic content. Organize your snippets into a library with tags, filters, and statuses so they’re easily findable.”

With Writer’s snippet shortcuts feature, users can call up a snippet anywhere they write with a shortcut, or search in-line for content. And with Writer’s admin functions, team leaders can set editorial style rules for punctuation and capitalization while enforcing reading grade level requirements across the organization.

Writer offers style guide management with templates and examples, allowing users to link to rules, term banks, and snippets libraries. The fonts, colors, and branding on style guide pages are customizable, and the pages can be published to the web for public consumption.

Competition

Enterprises are boosting their investments in tools like Writer that tap natural language processing (NLP), the subfield of linguistics and AI concerned with how algorithms analyze large amounts of language data. According to a survey from John Snow Labs and Gradient Flow, 60% of tech leaders indicated that their NLP budgets grew by at least 10% compared to 2020, while a third — 33% — said that their spending climbed by more than 30%.

Writer has competition in Grammarly, which similarly provides an AI-powered writing assistant for a range of use cases. But Writer asserts that Grammarly doesn’t boast the same style guide and “voice alignment” capabilities, nor features like gender-neutral pronoun and “plain language” conversion.

Writer counts Pinterest, Bill.com, Accenture, Deloitte, Twitter, and Intuit as customers. This year, annual recurring revenue tripled as the startup’s customer base reached 150 brands.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

Chip developer Cerebras bolsters AI-powered workload capabilities with $250M

Cerebras Systems, the California-based company that has built a “brain-scale” chip to power AI models with 120 trillion parameters, said today it has raised $250 million funding at a valuation of over $4 billion. Cerebras claims its technology significantly accelerates the time involved in today’s AI work processes at a fraction of the power and space. It also claims its innovations will support the multi-trillion parameter AI models of the future.

In a press release, the company stated that this additional capital will enable it to further expand its business globally and deploy its industry-leading CS-2 system to new customers, while continuing to bolster its leadership in AI compute.

Cerebras’ cofounder and CEO Andrew Feldman noted that the new funding will allow Cerebras to extend its leadership to new regions. Feldman believes this will aid the company’s mission to democratize AI and usher in what it calls “the next era of high-performance AI compute” — an era where the company claims its technology will help to solve today’s most urgent societal challenges across drug discovery, climate change, and much more.

Redefining AI-powered possibilities

“Cerebras Systems is redefining what is possible with AI and has demonstrated best in class performance in accelerating the pace of innovation across pharma and life sciences, scientific research, and several other fields,” said Rick Gerson, cofounder, chairman, and chief investment officer at Falcon Edge Capital and Alpha Wave.

“We are proud to partner with Andrew and the Cerebras team to support their mission of bringing high-performance AI compute to new markets and regions around the world,” he added.

Image of the Cerebras Wafer Scale Engine.

Cerebras’ CS-2 system, powered by the Wafer Scale Engine (WSE-2) — the largest chip ever made and the fastest AI processor to date — is purpose-built for AI work. Feldman told VentureBeat in an interview that in April of this year, the company more than doubled the capacity of the chip, bringing it up to 2.6 trillion transistors, 850,000 AI-optimized cores, 40GBs on-chip memory, 20PBs memory bandwidth, and 220 petabits fabric bandwidth. He noted that for AI work, big chips process information more quickly and produce answers in less time.

With only 54 billion transistors, the largest graphics processing unit pales in comparison to the WSE-2, which has 2.55 trillion more transistors. With 56 times more chip size, 123 times more AI-optimized cores, 1,000 times more high-performance on-chip memory, 12,733 times more memory bandwidth, and 45,833 times more fabric bandwidth than other graphic processing unit competitors, the WSE-2 makes the CS-2 system the fastest in the industry. The company says its software is easy to deploy, and enables customers to use existing models, tools, and flows without modification. It also allows customers to write new ML models in standard open source frameworks.

New customers

Cerebras says its CS-2 system is delivering a massive leap forward for customers across pharma and life sciences, oil and gas, defense, supercomputing centers, national labs, and other industries. The company announced new customers including Argonne National Laboratory, Lawrence Livermore National Laboratory, Pittsburgh Supercomputing Center (PSC) for its groundbreaking Neocortex AI supercomputer, EPCC, the supercomputing center at the University of Edinburgh, Tokyo Electron Devices, GlaxoSmithKline, and AstraZeneca.

A list of Cerebras's newest customers. The list of these customers can be found in the text of the article itself.

The series F investment round was spearheaded by Alpha Wave Ventures, a global growth stage Falcon Edge-Chimera partnership, along with Abu Dhabi Growth (ADG).

Alpha Wave Ventures and ADG join a group of strategic world-class investors including Altimeter Capital, Benchmark Capital, Coatue Management, Eclipse Ventures, Moore Strategic Ventures, and VY Capital. Cerebras has now expanded its frontiers beyond the U.S., with new offices in Tokyo, Japan, and Toronto, Canada. On the back of this funding, the company says it will keep up with its engineering work, expand its engineering force, and hunt for talents all over the world going into 2022.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

Microsoft acquires AI-powered moderation platform Two Hat

Microsoft today announced that it acquired Two Hat, an AI-powered content moderation platform, for an undisclosed amount. According to Xbox product services CVP Dave McCarthy, the purchase will combine the technology, research capabilities, teams, and cloud infrastructure of both companies to serve Two Hat’s existing and new customers and “multiple product and service experiences: at Microsoft.

“Working with the diverse and experienced team at Two Hat over the years, it has become clear that we are fully aligned with the core values inspired by the vision of founder, Chris c, to deliver a holistic approach for positive and thriving online communities,” McCarthy said in a blog post. “For the past few years, Microsoft and Two Hat have worked together to implement proactive moderation technology into gaming and non-gaming experiences to detect and remove harmful content before it ever reaches members of our communities.”

Moderation

According to the Pew Research Center, 4 in 10 Americans have personally experienced some form of online harassment. Moreover, 37% of U.S.-based internet users say they’ve been the target of severe attacks — including sexual harassment and stalking — based on their sexual orientation, religion, race, ethnicity, gender identity, or disability. Children, in particular, are the subject of online abuse, with one survey finding a 70% increase in cyberbullying on social media and gaming platforms during the pandemic.

Priebe founded Two Hat in 2012 when he left his position as a senior app security specialist at Disney Interactive, Disney’s game development division. A former lead developer on the safety and security team for Club Penguin, Priebe was driven by a desire to tackle the issues of cyberbullying and harassment on the social web.

Today, Two Hat claims its content moderation platform — which combines AI, linguistics, and “industry-leading management best practices” — classifies, filters, and escalates more than a trillion human interactions including messages, usernames, images, and videos a month. The company also works with Canadian law enforcement to train AI to detect new child exploitative material, such as content likely to be pornographic.

“With an emphasis on surfacing online harms including cyberbullying, abuse, hate speech, violent threats, and child exploitation, we enable clients across a variety of social networks across the globe to foster safe and healthy user experiences for all ages,” Two Hat writes on its website.

Microsoft partnership

Several years ago, Two Hat partnered with Microsoft’s Xbox team to apply its moderation technology to communities in Xbox, Minecraft, and MSN. Two Hat’s platform allows users to decide the content they’re comfortable seeing — and what they aren’t — which Priebe believes is a key differentiator compared with AI-powered moderation solutions like Sentropy and Jigsaw Labs’ Perspective API.

“We created one of the most adaptive, responsive, comprehensive community management solutions available and found exciting ways to combine the best technology with unique insights,” Priebe said in a press release. “As a result, we’re now entrusted with aiding online interactions for many of the world’s largest communities.”

It’s worth noting that semi-automated moderation remains an unsolved challenge. Last year, researchers showed that Perceive, a tool developed by Google and its subsidiary Jigsaw, often classified online comments written in the African American vernacular as toxic. A separate study revealed that bad grammar and awkward spelling — like “Ihateyou love,” instead of “I hate you,” — make toxic content far more difficult for AI and machine detectors to spot.

As evidenced by competitions like the Fake News Challenge and Facebook’s Hateful Memes Challenge, machine learning algorithms also still struggle to gain a holistic understanding of words in context. Revealingly, Facebook admitted that it hasn’t been able to train a model to find new instances of a specific category of disinformation: misleading news about COVID-19. And Instagram’s automated moderation system once disabled Black members 50% more often than white users.

But McCarthy expressed confidence in the power of Two Hat’s product, which includes a user reputation system, supports 20 languages, and can automatically suspend, ban, and mute potentially abusive members of communities.

“We understand the complex challenges organizations face today when striving to effectively moderate online communities. In our ever-changing digital world, there is an urgent need for moderation solutions that can manage online content in an effective and scalable way,” he said. “We’ve witnessed the impact they’ve had within Xbox, and we are thrilled that this acquisition will further accelerate our first-party content moderation solutions across gaming, within a broad range of Microsoft consumer services, and to build greater opportunity for our third-party partners and Two Hat’s existing clients’ use of these solutions.”

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

Cisco launches new AI-powered and hybrid event features for Webex

Join gaming leaders online at GamesBeat Summit Next this upcoming November 9-10. Learn more about what comes next. 


Cisco today unveiled updates across its Webex portfolio of communications products, including an integrated asynchronous camera feature, AI-powered sound, video enhancements, and a management service for hybrid in-person and virtual events. The company’s upgrades are designed to power events and meetings “at scale” while maintaining interoperability with Zoom, Microsoft Teams, Google Meet, and other third-party meeting platforms.

Headwinds from the pandemic have driven the value of the global videoconferencing market in 2021 to an estimated $6.03 billion. Sixty percent of respondents to an Owl Labs survey reported participating in video meetings more often in 2020 than in 2019, a bigger rise than other workplace staples like email saw year-over-year. Dovetailing with this, hybrid events are expected to continue to have a presence in work life, with 79% of companies expecting to host events that include a virtual attendance option, according to Martech.

“Technology has many powers, and the greatest of all is its ability to connect people and level the playing field for so many across the globe,” Cisco security and collaboration executive VP and general manager Jeetu Patel said in a statement. “Our new Webex innovations mark a significant step forward in helping our customers unlock the potential of their hybrid workforce — enabling them to collaborate in new ways and drive [an] inclusive experience.”

AI and hybrid work

Cisco is rolling out “AI-powered audio intelligence” in Webex, leveraging an AI algorithm to optimize all participants’ voices during meetings. The setting equalizes voices regardless of how close they’re to their devices, automatically differentiating intended speech from background noise.

Another AI-powered feature, People Focus, will be available in December. It will provide “better clarity” and “optimized visuals” of in-room attendees’ facial gestures and body language, Cisco says. Additional camera-related enhancements coming in early 2022 will further improve the interface in meeting rooms, including showing conference room participants in individual boxes on-screen — regardless of which meeting platform they use.

In related news, Webex Assistant, Cisco’s virtual meeting tool, now supports French, German, Spanish, and Japanese in addition to English. In August, it gained the ability to translate closed captions from English into more than 100 languages with a paid add-on. And starting this week, developers can work with partners to design custom commands for Webex Assistant running on Cisco’s Webex devices such as desktop cameras, headsets, and conference room phones.

Vidcast, an asynchronous communication service, also joins the list of new Webex features. Currently in beta at Vidcast.io ahead of integration with the Webex App in Spring 2022, it provides the ability to record, watch, comment, and react to meeting clips on-demand.

Meanwhile, Webex’s new Whiteboarding tool enables users to create, find, edit, and share whiteboards with anybody, not excluding people outside their organization. Users can annotate using any device — mobile, tablet, laptop — in addition to Webex devices.

Webex also now features Collaboration Insights, offering personalized details like the top ten people a user collaborates with weekly, new colleague titles, and more. Two complementary capabilities — Well-being and Cohesion — in the previously announced People Insights tab give teams a view into anonymous work time patterns, sentiment ratings, and focus time goals. Exclusively for Webex Suite customers, there’s Thrive Reset, a collection of wellness topics, and a gallery where users can upload photos. It’s based on research showing that it takes 60 to 90 seconds to course-correct from stress, Cisco says, and designed to help users “focus on breathing, reflect on what they’re grateful for, reframe problems, or simply stand up and stretch.”

“[W]hen we provide insights … to an individual, the individual owns the data, not the organization because we don’t believe that without your explicit permission, you’d want to have your boss see that,” Patel told VentureBeat in a previous interview regarding Webex’s new monitoring features. “Engagement should not be measured based on having a judgment on someone saying, ‘I’m judging that you look sad, and therefore I’m going to do certain things’ … at that point in time, in my mind, you could cross a boundary where there’s more bad that can come out of that than good … There’s a fine line between ‘This is super productive’ and ‘We can’t do this because it violates my privacy, or it’s just outright creepy.’”

Events, integrations, and devices

Following Cisco’s acquisitions of Socio Labs and Slido earlier this year, the company unveiled an expanded Webex Events product targeting enterprises hosting hybrid events. Management capabilities spanning badging and printing for ticketing, monetization, and networking are available, and customers can now host events via Webex Webinars with Slido’s polling, quizzing, and Q&A technology up to 10,000 attendees (in webinar mode) or 100,000 (in webcast mode) in size.

Today, Cisco also announced its 60-plus new partner integrations to Webex including Smartsheet, Hacker Rank, Thrive Reset, Miro, and Mural.

Against this backdrop, new Webex devices are coming to market — among them the Webex Desk Mini. The Webex Desk Mini, which comes in a range of colors, features a 15.6-inch, 1080p interactive display; a 64-degree HD camera; a full-range speaker; and a background noise removal mic array. Meanwhile, the new Webex Board Pro sports dual 4K cameras, directional audio, two active styluses, and a choice of a 55- or 75-inch display.

Webex Desk Mini will be available to order in early 2022 for $1,695. Existing Webex enterprise customers will receive the “cloud promo” price of $999. The Webex Board Pro will launch in available in November, priced at $11,995 (55 inches) and $19,995 (75 inches).

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link

Categories
AI

AI-powered spend management platform Yokoy secures $26M

Spend management platform Yokoy today announced that it raised $26 million in series A financing from Left Lane, with participation from prominent European investor Balderton Capital. The Zurich, Switzerland-based company says that it’ll use the new funding — which brings its total raised to $27.1 million — to expand across the U.S., Europe, and other regions around the world while enhancing the technologies underpinning its platform.

The spend management software industry is projected to be worth $3.97 billion by 2027, up from $1.08 billion in 2019, according to Verified Market Research. The Global Business Travel Association pegs the cost to companies of processing a single expense report at $58 each time, while processing a single invoice is estimated to cost between $12 and $32. Indeed, the biggest cost drivers when it comes to spend management are often errors, manual work, and transaction fees. Yokoy asserts that these costs can be cut by half if certain resource-intensive processes are automated.

Founded in Switzerland in 2019, Yokoy provides an AI-powered spend management suite for midsize and enterprise companies. Through a combination of machine learning, automation, and API integrations, the startup offers expense management, supplier invoice management, and corporate credit cards.

CEO Philippe Sahli, and chief technology officer Devis Lussi (who previously worked at CERN) met while working at Ernst & Young’s management consultancy, while chief customer officer Lars Mangelsdorf had been building software-as-a-service (SaaS) products at other startups. Meanwhile, CFO Thomas Inhelder — who’s held accounting roles at KPMG — came into contact with Sahli at a previous startup.

“With Yokoy, we’re building a highly intelligent, highly secure, and highly customizable global spend management platform that empowers our customers to take control of their vast corporate spending processes and fine-tune their workflows,” Lussi said in a statement. “We’re helping them to cut costs, save time and bring clarity to their global operations in a way that fits their ambitions. It’s this that we believe will see us becoming the leading spend management platform in the world.”

AI-powered spend management

With Yokoy’s platform, midsize to enterprise companies can configure and build their own process flows and integrate Yokoy with third-party tools. Lussi claims that the platform is “self-learning,” enabling Yokoy to monitor individual workflows and processes to make them more efficient and scalable over time.

For example, customers can import expense receipts and invoices by snapping pictures of them through Yokoy’s mobile app. The platform’s algorithms enhance the picture quality before extracting the words and numbers via optical character recognition, validating more than 300 data points in a single receipt or invoice. In the case of invoices, Yokoy can also recognize suppliers, match them with data from a company’s enterprise resource planning software, and fill any missing data into the scanned document. With the information it extracts from documents, Yokoy checks relevant policies, gauges the potential for fraud, and validates the data for outliers and rules violations.

Yokoy

Above: Yokoy’s spend management platform.

Image Credit: Yokoy

Any scans that don’t pass Yokoy’s quality assurance benchmarks are set aside for manual review, while the rest are automatically exported to an accounting system.

“Yokoy is able to reconstruct [the] context [of documents] based on various features. Such features can be the relative position of a sequence of characters on the paper — top, bottom, left, right — or the presence of certain keywords. With the help of keywords (Yokoy’s list comprises more than 100,000 entries in many different languages), something can be found out about the type of an expense,” Sahli told VentureBeat via email. “Yokoy has described the relationships between these features within [AI] models. In the fraction of a second [that] it takes the software to digitize and analyze an expense receipt, several such models are processed. The models have been trained with millions of examples, and they are constantly being refined.”

Yokoy competes with Ramp and others in the spend management solutions segment. But in two years, the company has managed to attract 80,000 users across 400 customers including DPD Group, Stadler Rail, Russia’s Sberbank, the Swiss bank Swissquote. Part of the latest investment will be put toward growing the company’s headcount by the end of 2021, Sahli says.

“Spend management includes the processing of supplier invoices, travel expenses, corporate credit card expenses, and all such accounts payable categories. While the travel expenses have decreased during the pandemic, other categories — like online purchases over corporate credit cards — have increased during that time,” Sahli added. “The total spend volume processed over Yokoy per customer during the pandemic has actually increased. Yokoy is a true winner of the pandemic, but for a completely different reason: COVID-19 has truly accelerated the digitization and automation in companies and that’s what has pushed us, especially the new customers number.”

Yokoy employs about 100 people throughout its five offices. It expects to more than double that number to 250 by the end of 2022.

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link