AI

Terzo lands $16M to extract key data from contracts • ZebethMedia

By zebethcontrolPosted on November 17, 2022Posted in AI, Enterprise, funding, Robotics & AI, startup, StartupsNo Comments

Contract governance is the steps taken to make sure agreed-upon terms between a company and its suppliers are met. It’s an essential part of doing business, and the consequences for getting it wrong can be steep. McKinsey estimates that poor contract governance can cost organizations up to 9% of their total revenue, which equates to $1.4 trillion for the Fortune 500 alone and $6.4 trillion across all enterprise business-to-business companies. Challenges around contract governance have fueled the rise of startups like Icertis, which recently secured $150 million at a $3.2 billion valuation to build out its contracting tools. LinkSquares in April landed $100 million for its AI-powered contract analysis platform, while ContractPodAi, a close competitor, has raised tens of millions to digitize contract reviews. A relatively new entrant in the space is Terzo, which was co-founded by Brandon Card, Al Giocondi and Pradeep Thangavel in 2020. A suite of contract processing software, Terzo uses AI to extract data in contracts related to a company’s spend and revenue across their supplier and customer relationships. In a sign investor interest in contract management startups hasn’t waned, Terzo today closed a $16 million Series A round led by Align Ventures with participation from TYH Ventures, Engage Ventures, Human Capital and other unnamed institutional investors. The proceeds bring the company’s total raised to more than $18f million, and Card, who serves at Terzo’s CEO, says they’ll be put toward Terzo’s sales and marketing initiatives as well as enhancing the platform’s AI capabilities. “As our technology evolves, we aim to deliver advanced insights around financials and budgeting,” Card said. “Contract systems were built for legal use cases and legal teams to focus on drafting and clauses. There are no analytics or financial insights for leaders to make smarter decisions. Terzo was founded to solve that problem.” Image Credits: Terzo Card says his experiences at Microsoft, where he was an enterprise portfolio manager at Microsoft Cloud, inspired him to co-found Terzo. Giocondi came from account manager roles at Oracle and IBM. As for Thangavel, he spent nearly seven years on the engineering side at Freshworks prior to joining alongside Card and Giocondi. Card says that Terzo’s AI was trained using real-world business contracts to extract data such as inventory and costs, supervised by a quality assurance team to ensure baseline accuracy. Terzo integrates with enterprise resource platforms like SAP and Oracle to track contractual obligations and expiration dates, including metrics related to environmental, social and governance policies. “Terzo is valuable to the IT audience because it allows them to see data faster,” Card said. “We’ve created a platform that combines data management, automation and AI, but also keeps people in the loop.” VCs see promise in contract management legal tech like Terzo’s, perhaps in part because of the high customer adoption rate. According to a 2020 Bloomberg Law survey, more than half (56%) of in-house lawyers said that they’re using contract management programs. (That’s despite the fact that the legal industry is notoriously slow to adopt new tech.) If the current trend holds, Markets and Markets predicts the contract management life cycle market will grow from $1.5 billion in 2019 to $2.9 billion by 2024. Card says that Terzo has “over a dozen” customers, including a Fortune 50 retailer and the largest financial transaction processor in the world. The plan into the next year is to drive revenue at over 50% of Terzo’s overall expenses, he says, and to break even in 2024. “We are positioned to be a profitable business by 2025,” Card added. “Both the pandemic and current tech slowdown has taught us how to run a lean business that is focused on efficient growth. The current downturn has caused our customers to prioritize spending and budgeting so we look at this as a positive tailwind heading into 2023.”

Namecoach raises cash to teach users how to correctly pronounce names • ZebethMedia

By zebethcontrolPosted on November 17, 2022Posted in AI, Enterprise, funding, Robotics & AI, startup, StartupsNo Comments

We’ve all been faced with a name that’s difficult to pronounce. But not everyone considers the consequences of their mispronunciations. In a piece for Fast Company, Madhumita Mallick, the head of inclusion, equity and impact at Carta, recalls how her name became a source of anxiety even when she was a grade-school student. Studies have indeed found that when peers and teachers incorrectly pronounce or change the names of students of color, those students participate less in class and becoming socially withdrawn to avoid associating with their name. As the co-authors of one wrote in 2012: “actions and attitudes [students] experienced in K-12 schools highlight a type of cultural ‘othering’ that contradicts our goals for multicultural school environments .. even just stumbling over a name they had never seen before, the tone set by a teacher about a student’s name [is] something significant.” Praveen Shanbhag, the CEO of Namecoach, heard a speaker at his sister’s alma mater flub her name during her college graduation. That, along with his experiences as a first-generation immigrant to the U.S., inspired him to co-found Namecoach, which develops name pronunciation tools that can be embedded in existing platforms like Salesforce, Canvas and Gmail. “We are on a mission to make name mispronunciation a thing of the past for everyone,” Shanbhag told ZebethMedia via email. Shanbhag started dabbling in programming while working on his Ph.D. in philosophy at Stanford. He decided to code an app that would collect recordings of students saying their names and deliver them to name readers for graduation, which became Namecoach. By 2016, Shanbhag says that hundreds of schools were using Namecoach’s software and services. Namecoach — which eventually broadened its customer base to client brands — isn’t alone in the market. NameShouts offers a robust set of name pronunciation tools, as does Facebook, Slack and LinkedIn. But Shanbhag tells me that early on, he sought to differentiate Namecoach by investing heavily in AI and integrations. Image Credits: Namecoach For example, Namecoach uses an AI system to predict the correct pronunciation when someone’s name has multiple correct pronunciations based on factors like nationality, ethnicity, gender and location. Another of the platform’s systems synthesizes speech in situations where an audio recording of a name isn’t available. At a high level, Namecoach shows name pronunciations both user-generated and drawn from a database of audio name pronunciations. When asked about the system’s accuracy and whether users can remedy mistakes that might make their way onto the database, Shanbhag said that Namecoach consults with linguistic experts and offers a submission form for corrections. It’s not yet live, but Shanbhag says that Namecoach is developing a AI to provide pronunciations that take the speaker’s native language into account — not just the name-owner’s language. “Your name is central to your identity, and accurate pronunciation sets the tone for a positive interaction for both parties,” he said. Namecoach’s platform works out-of-the-box with services including Microsoft Teams, Outlook and Google Workspace and offers an API and software development kit to let third parties build Namecoach’s functionality into their products. But the focus over the next few months will be the startup’s first-ever consumer app, Shanbhag says, which will be made available as a Chrome extension next year. Namecoach — which today closed an $8 million Series A round led by Impact America Fund with participation from Authentic Ventures, Metaplanet, Engage.VC, Founders Fund, Bisk Ventures and others — also plans to ramp up its sales and marketing efforts and expand the platform’s collection of connectors. Shanbhag hinted at capabilities beyond pronunciation guidance coming down the pipeline, including prompts to enable warm interactions in sales scenarios and feedback to help users improve their conversations more generally. Leaning into sales applications makes sense given Namecoach’s marquee customers — Salesforce, Netjets and PwC. Beyond those three, the 30-person startup claims to have more than 300 education and corporate clients worldwide. “The Series A funding enables us to accelerate our goal of integrating novel voice technology into every communication workflow to solve name pronunciation across a wide spectrum of use cases during any voice interaction,” said Shanbhag, who wouldn’t Namecoach’s disclose revenue figures when asked. “Namecoach has customers across a very wide range of verticals, use cases and organization size, which means that we are not reliant on any subsegment to thrive in a downturn.” To date, Palo Alto-based Namecoach has raised $15 million in venture capital.

Microsoft and Nvidia team up to build new Azure-hosted AI supercomputer • ZebethMedia

By zebethcontrolPosted on November 16, 2022Posted in AI, Enterprise, Microsoft, nvidia, Robotics & AI, supercomputerNo Comments

Roughly two years ago, Microsoft announced a partnership with OpenAI, the AI lab with which it has a close commercial relationship, to build what the tech giant called an “AI Supercomputer” running in the Azure cloud. Containing over 285,000 processor cores and 10,000 graphics cards, Microsoft claimed at the time that it was one of the largest supercomputer clusters in the world. Now, presumably to support even more ambitious AI workloads, Microsoft says it’s signed a “multi-year” deal with Nvidia to build a new supercomputer hosted in Azure and powered by Nvidia’s GPUs, networking and AI software for training AI systems. “AI is fueling the next wave of automation across enterprises and industrial computing, enabling organizations to do more with less as they navigate economic uncertainties,” Scott Guthrie, executive vice president of Microsoft’s cloud and AI group, said in a statement. “Our collaboration with Nvidia unlocks the world’s most scalable supercomputer platform, which delivers state-of-the-art AI capabilities for every enterprise on Microsoft Azure.” Details were hard to come by at press time. But in a blog post, Microsoft and Nvidia said that the upcoming supercomputer will feature hardware like Nvidia’s Quantum-2 400Gb/s InfiniBand networking technology and recently-detailed H100 GPUs. Current Azure instances offer previous-gen Nvidia A100 GPUs paired with Quantum 200Gb/s InfiniBand networking. Notably, the H100 — the flagship of Nvidia’s Hopper architecture — ships with a special “Transformer Engine” to accelerate machine learning tasks and — at least according to Nvidia — delivers between 1.5 and 6 times better performance than the A100. It’s also less power-hungry, offering the same performance as the A100 with up to 3.5 times better energy efficiency. One of the first industrial-scale machines to sport H100 GPUs, the Lenovo-built Henri system operated by the Flatiron Institute in New York City, topped the list of this year’s most efficient supercomputers. As part of the Microsoft collaboration, Nvidia says that it’ll use Azure virtual machine instances to research advances in generative AI, or the self-learning algorithms that can create text, code, images, video or audio. (Think along the lines of OpenAI’s text-generating GPT-3 and image-producing DALL-E 2.) Meanwhile, Microsoft will optimize its DeepSpeed library for new Nvidia hardware, aiming to reduce computing power and memory usage during AI training workloads, and work with Nvidia to make the company’s stack of AI workflows and software development kits available to Azure enterprise customers. Why Nvidia would opt to use Azure instances over its own in-house supercomputer, Selene, isn’t entirely clear; the company’s already tapped Selence to train generative AI like GauGAN2, a text-to-image generation model that creates art from basic sketches. Evidently, Nvidia anticipates that the scope of the AI systems that it’s working with will eventually surpass Selene’s capabilities. “AI technology advances as well as industry adoption are accelerating. The breakthrough of foundation models has triggered a tidal wave of research, fostered new startups and enabled new enterprise applications,” Manuvir Das, VP of enterprise computing at Nvidia, said in a statement. “Our collaboration with Microsoft will provide researchers and companies with state-of-the-art AI infrastructure and software to capitalize on the transformative power of AI.” The insatiable demand for powerful AI training infrastructure has led to an arms race of sorts among cloud and hardware vendors. Just this week, Cerabras, which has raised over $720 million in venture capital to date at an over-$4 billion valuation, unveiled a 13.5-million core AI supercomputer called Andromeda it claims can achieve more than 1 exaflop of AI compute. Google and Amazon continue to invest in their own proprietary solutions, offering custom-designed chips — e.g. TPUs and Trainium — for accelerating AI training in the cloud. The push for more powerful hardware will continue for the foreseeable future. A recent study found that the compute requirements for large-scale AI models has been doubling at an average rate of 10.7 months between 2016 and 2022. And OpenAI once estimated that, if GPT-3 were to be trained on a single Nvidia Tesla V100 GPU, it would take around 355 years.

TheGist taps AI to summarize Slack channels and threads • ZebethMedia

By zebethcontrolPosted on November 15, 2022Posted in AI, aleph, App, Apps, Enterprise, Robotics & AI, slack, summarization, TheGistNo Comments

Itay Dressler and Itzik Ben Bassat, who’ve held various software engineering and executive roles at startups together over the years, are accustomed to exchanging brief messages. Ben Bassat has ADHD, and for that reason prefers to keep texts on the shorter side. But as he and Dressler were faced with wrangling an increasing number of tools at their employers, they came to realize they weren’t the only ones who could benefit from more succinct updates. So they founded TheGist with the grand mission of “simplifying information consumption in workplace communications and data” through instant highlights. The startup’s first product uses AI to scan Slack messages and provide a personalized summary, aiming to filter out noise. And in the enterprise, there’s plenty of noise to filter. According to a 2021 report in Tech Republic, a survey of remote workers showed that 18% suffered from “information overload” while 8% were overwhelmed by the amount of data and apps they were meant to check each day. “There’s an overload of software-as-a-service (SaaS) applications that aren’t deeply integrated. Different teams use different tools to create information silos,” Ben Bassat told ZebethMedia in an email interview. “The integration between those SaaS tools makes the information overload greater, not smaller. There is no reason that in 2022, using AI, employees can’t get the information they need to make better decisions in a short and personalized form.” Installing TheGist’s Slack app — which can summarize both channels and threads — is a straightforward-enough process. Once connected to a workspace, the app can be added or invited to channels that a user wishes to summarize. Typing the command “/gist” summons it, generating a fresh summary — generally a bullet point or two in length — of what happened in the channel, visible only to the person who requested it. Image Credits: TheGist TheGist Slack app can provide summaries covering time scales from one day to several weeks. Beyond this, it can summarize particularly long individual Slack messages. Service is free for up to five summaries but unlimited summaries requires a premium subscription, which starts at $10 per user per month. “We wanted to release a tool that highlights the need for shortening the information overload in companies,” Ben Bassat said. “TheGist is a game changer for decision makers as we enable managers to dramatically increase the amount of workplace information they can consume by digesting it and personalizing it … For employees, we serve them the information they need when they need it so they can be aligned with the organization and make better and more knowledgeable decisions.” That’s a lot to promise. AI, while improving by leaps and bounds, has its limitations; TheGist’s summaries are bound to contain mistakes from time to time. And from a security standpoint, companies might be loathe to let a third-party app process the internal messages — particularly companies in highly regulated industries. Ben Bassat didn’t provide much in the way of detail around TheGist’s AI systems and their development, save that it’s leveraging “multiple open source large languages models” with “specific in-house fine-tuning.” “We are using statistical models to evaluate our models’ output and assess correctness,” Ben Bassat said. “As in every product which is generated by AI, results can have summary errors, and our users are made aware of that.” On the compliance question, Ben Bassat claims that TheGist doesn’t store Slack data other than the specific messages users ask to summarize, which it deletes after the summaries are generated. “We only store analytical and usage data in order to improve our product and personalize the user experience. Users can ask to delete their data according to our privacy policy,” Ben Bassat added. There aren’t a lot of competitors in the Slack summarization space. But there are a few, it’s worth noting. Frame summarizes the previous day’s Slack activity, providing metrics including team responsiveness and auto-detected “high” and “low” moments. Grok, a Slack app, provides summaries of Slack conversations and threads generated by OpenAI’s GPT-3 API. There’s also TLDR, which uses algorithms to spit out Slack message summaries. Image Credits: TheGist But Ben Bassat and co don’t see TheGist’s first app as the endgame. In parallel to it, Ben Bassat says that the company’s on the cusp of releasing “proprietary generative AI solutions” for different platforms in the near future — although it’s not clear for which platforms and what types of generative AI. Ben Bassat didn’t have much to say on the subject, which suggests that the specifics are in flux. “The goal of our platform is to enable anyone to be informed with short updates from any app they use for communication or productivity: Email, texts, project managing tools, doc files and more. Solving this challenge requires a lot of technological focus with a high level of expertise,” Ben Bassat said. “Our vision is to provide accurate summaries and actionable insights across all information-producing apps.” The success of TheGist’s Slack app aside, generative AI is probably a wise path to take. It’s the hot new thing in tech, to be sure, with startups like Jasper, an AI copywriting app for marketers, recently raising $125 million at a $1.5 billion valuation. VCs are certainly excited by the prospect; Sequoia Capital said in a blog post from September that it thought generative AI could “create trillions of dollars of economic value.” For its part, TheGist has raised $7 million to date in pre-seed funding co-led by StageOne Ventures and Aleph. Eden Shochat, a partner at Aleph, said via email: “TheGist’s debut tool is only the starting point, and there is so much more to come. In a world where companies create excessive amounts of data, across multiple tools, employees only want to zero in on the insights that matter to them, at the point in time when they are relevant. TheGist is on a mission to create magical tools that work for the user, rather than the other way around.”

Modus expands to sub-Saharan Africa with the launch of its AI and blockchain-focused $75M fund • ZebethMedia

By zebethcontrolPosted on November 15, 2022Posted in africa, AI, blockchain, modus, Startups, VentureNo Comments

New York-based venture platform Modus has launched Modus Africa, a venture capital fund for AI and blockchain startups across sub-Saharan Africa, ZebethMedia has learned. The fund is expected to reach a final close in the first quarter of next year. The spinoff continues Modus’s string of moves over the past 18 months, which has seen it add branches in Abu Dhabi, Cairo, and, most recently, Riyadh, supported by institutions like Mubadala’s Hub71. Modus says that its entry into Africa creates an “additional conduit of market access for Modus portfolio companies while also enabling African startups to scale into the MENA region.” As a “holistic venture platform,” Modus runs three business units focusing on entrepreneurs and startups in the MENA and GCC regions. They include the Venture Builder, which works with idea and early-stage MVP stage companies. Then there’s Corporate Innovation, a service platform that leverages the firm’s internal know-how to support corporations and government entities. And its Venture Capital arm provides investment to early and midstage-sized startups, such as staffing platform Ogram. On its website, Modus says its fund is backed by several investors ranging from UHNWI, family offices, private investors, and government-backed entities from the U.S., the EU, and MENA. Although Modus primarily invests in foreign-based companies that are “portable to the Middle East,” as well as startups in Egypt and the GCC, its expansion into sub-Saharan Africa isn’t surprising. Last year, African startups raised over $5 billion and minted five unicorns (per this report, the continent observed a 250% year-over-year growth in funding and surpassed capital deployed in MENA). And despite the current macroeconomic trends and conditions that have resulted in layoffs, down rounds and shutdowns, startups on the continent are set to top last year’s fundraising record numbers. Unlike other firms with marked funds interested in Africa, Modus’s interest in AI and blockchain technologies is intriguing. Though it has household names such as Tunisia’s InstaDeep, Kenya’s Sama, and South Africa’s DataProphet — and several web3 startups claiming to build on the blockchain — Africa’s AI and blockchain sectors are still relatively nascent. The thinking behind adopting this strategy can be traced to Vianney Mathonnet and Andre Jr. Ayotte, the general partners of Modus’s Africa-focused fund. Both partners, in an interview with ZebethMedia, described how several stints working in banking, finance and Dubai-based family offices pointed them to the emergence of blockchain technology and its outsized opportunity and application in Africa. “Not long after we launched this project after noticing how massive blockchain and AI could be in Africa, we were approached by Modus Capital because they wanted a Pan-African strategy themselves,” said Ayotte. “They were looking for people with the know-how, the network, experience to do that, so we started discussing how the partnership would work. Ultimately, what happened is that our project became the Modus Africa fund.” L-R: Andre Jr. Ayotte and Vianney Mathonnet (General Partners, Modus Africa) According to a statement, Modus says Africa has the potential of reaching 200 million+ new blockchain users in the next four years, fueled by necessity and a fast-growing tech-savvy population. Nevertheless, the six-year-old VC firm isn’t only taking a chance on purely AI and blockchain startups; instead, it is cutting checks in startups across broader sectors that are implementing those technologies into their products. The firm is currently closing three investments in startups using AI and blockchain across insurtech, fintech and health tech, said the general partners who control the fund’s thesis, direction and investment strategy while leveraging Modus’s 50+ team to carry out due diligence and portfolio management. Mathonnet said the “jurisdiction-agnostic” Modus Africa will invest in about 45 seed-stage startups and allocate 50% of the $75 million SDG-focused fund for follow-on investments, especially in Series A rounds. These checks will range from $350,000 to $1.2 million across both stages. “We as a fund will reinvest in our winners and our LPs are also looking to reinvest in them outside the fund, catalyzing even more money in the ecosystem in Africa,” said the partners. “In terms of countries, we know that tech talent and incubators are really strong in tech ecosystems like Kenya and Nigeria, Egypt, and South Africa, and it’s inevitable that a good deal flow within all these regions. With that said, though, we are exploring new regions and searching for key partnerships to enter those markets and add some support and sustainability for deal flow.” Some of these markets include the Democratic Republic of Congo (DRC), Niger and others in Francophone Africa. Speaking on the formation of Modus Africa, Kareem Elsirafy, the managing partner of Modus, said in a statement: “Modus is proud to be launching an Africa-MENA investment corridor to continue supporting and investing in emerging innovation ecosystems. The Modus platform is uniquely positioned to deliver impact and value to African communities through operational, institutional, and financial capital. We’re excited to have Vianney and Andre leading the way on this journey.”

DeviantArt provides a way for artists to opt out of AI art generators • ZebethMedia

By zebethcontrolPosted on November 11, 2022Posted in AI, Apps, deviantART, Media & Entertainment, Robotics & AI, text-to-imageNo Comments

DeviantArt, the Wix-owned artist community, today announced a new protection for creators to disallow art-generating AI systems from being developed using their artwork. An option on the site will allow artists to preclude third parties from scraping their content for AI development purposes, aiming to prevent work from being swept up without artists’ knowledge or permission. “AI technology for creation is a powerful force we can’t ignore. . . . It would be impossible for DeviantArt to try to block or censor this art technology,” CEO Moti Levy told ZebethMedia in an email interview. “We see so many instances where AI tools help artists’ creativity, allowing them to express themselves in ways they could not in the past. That said, we believe we have a responsibility to all creators. To support AI art, we must also implement fair tools and add protections in this domain.” As AI-generated artwork began to proliferate on the web earlier this year, fueled by the release of text-to-image tools like Stable Diffusion and DALL-E 2, art-housing platforms were forced to take a policy stance. Some, including Newgrounds, PurplePort and Getty Images, banned AI-generated art altogether, concerned both about the impact to artists and the legal ramifications of art created by tools that were developed on copyrighted works. Today’s bleeding-edge AI art tools “learn” to generate new images from text prompts by “training” on billions of existing images, which often come from data sets that were scraped together by trawling public image hosting websites like Flickr and ArtStation. Some legal experts suggest that training AI models by scraping public images — even copyrighted ones — will likely be covered by fair use doctrine in the U.S. But it’s a matter that’s unlikely to be settled anytime soon — particularly in light of contrasting laws being proposed overseas. OpenAI, the company behind DALL-E 2, took the proactive step of licensing a portion of the images in DALL-E 2’s training data set. But the license was limited in scope, and rivals so far haven’t followed suit. “Many creators are rightfully critical of AI-generation models and tools. For one, they do not give creators control over how their art may be used to train models, nor do they let creators decide if they authorize their style to be used as inspiration in generating images,” Levy continued. “As a result, many creators have seen AI models being trained with their art or worse: AI art being generated in their style without the ability to opt out or receive proper credit.” Art created with DeviantArt’s DreamUp tool. Image Credits: Digitonaut / DeviantArt DeviantArt’s new protection will rely on an HTML tag to prohibit the software robots that crawl pages for images from downloading those images for training sets. Artists who specify that their content can’t be used for AI system development will have “noai” and “noimageai” directives appended to the HTML page associated with their art. In order to remain in compliance with DeviantArt’s updated terms of service, third parties using DeviantArt-sourced content for AI training will have to ensure that their data sets exclude content that has the tags present, Levy says. “DeviantArt expects all users accessing our service or the DeviantArt site to respect creators’ choices about the acceptable use of their content, including for AI purposes,” Levy added. “When a DeviantArt user doesn’t consent to third party use of their content for AI purposes, other users of the service and third parties accessing the DeviantArt site are prohibited from using such content to train an AI system, as input into any previously trained AI system or to make available any derivative copy unless usage of that copy is subject to conditions at least as restrictive as those set out in the DeviantArt terms of service.” It’s an attempt to give power back to artists like Greg Rutkowski, whose classical painting styles and fantasy landscapes have become one of the most commonly used prompts in the AI art generator Stable Diffusion — much to his chagrin. Rutkowski and others have expressed concern that AI-generated art imitating their styles will crowd out their original works, harming their income as people start using AI-generated images for commercial purposes. The tools have set off firestorms of controversy in recent months. A system trained to imitate the style of acclaimed South Korean illustrator Kim Jung Gi, who passed away suddenly in early October, was condemned by many in the art community as a tasteless stunt. After winning a prize at the Colorado State Fair’s art competition, artwork made by AI set off a fierce backlash. Elsewhere, character designers like Hollie Mengert have decried what they see as poor AI imitations of their style that are nevertheless inexorably tied to their names. For DeviantArt’s part, it’s encouraging creator platforms to adopt artist protections and says it’s already in discussions about implementation with “several players.” But it’s unclear whether it’ll be able to rally the broader industry behind its approach; less scrupulous actors could theoretically ignore DeviantArt’s terms of service to scrape images regardless of HTML tag. Technologists Mat Dryhurst and Holly Herndon are spearheading a separate effort called Source+ to let people disallow their work or likeness to be used for AI training purposes. Meanwhile, Shutterstock is banning all AI art not created with DALL-E 2 to mitigate copyright issues (and likely to preserve its partnership with OpenAI). Image Credits: Digitonaut / DeviantArt Unlike Shutterstock, DeviantArt has allowed — and will continue to allow — art generated with third-party AI tools on its platform, Levy says, though it encourages users uploading AI-generated art to tag it as such. He claims that tens of thousands of images tagged as “AI-art” are being submitted to DeviantArt each month, growing over 1,000% in the last four months. “Since DeviantArt’s inception, we’ve never believed in blocking any art genres or categories. We have always made room for and supported all types of creators and their works,” Levy said. Beyond simply allowing AI art, DeviantArt is committing to

OpenAI leads $23.5M round in Mem, an AI-powered note-taking app • ZebethMedia

By zebethcontrolPosted on November 10, 2022Posted in AI, Enterprise, funding, Mem, note taking, OpenAI, Robotics & AI, StartupsNo Comments

Last year, OpenAI announced the OpenAI Startup Fund, a tranche through which it and its partners, including Microsoft, are investing in early-stage AI companies tackling major problems. Mum’s been the word since on which companies have received infusions from the Fund. But today, the OpenAI Startup Fund revealed that it led a $23.5 million investment in Mem, a work-focused app that taps AI to automatically organize notes. The investment values Mem at $110 million post-money and brings the startup’s total raised to $29 million. Co-founded by Kevin Moody and Dennis Xu, Mem differentiates itself from traditional note-taking apps by emphasizing “lightweight organization,” in Moody and Xu’s words. The workflow revolves around search and a chronological timeline, allowing users to attach topic tags, tag other users and add recurring reminders to notes. Mem users can capture quick notes, send links and save images from anywhere using SMS, messaging apps and the platform’s mobile client. Collaboration features let teams share, edit and comment on notes and directly attach them to shared calendars for faster reference. Mem’s search experience uses AI to search across notes, aiming to understand which notes might be most relevant in a given moment to a particular person. Moody and Xu say the platform is designed to augment knowledge workers in their typical responsibilities, like reading through pages of information, extracting the pieces relevant to a particular question and transforming the information into an answer or a report. Mem taps AI to organize notes in real time. There’s no doubt knowledge-seeking tasks are time-consuming. According to Gartner, professionals spend 50% of their working hours searching for information and on average take 18 minutes to locate a file (albeit the veracity of metrics like these has been challenged over the years). One source estimates that document disorganization costs businesses $3,900 per employee each year in productivity losses, making Mem an attractive proposition if the tech works as advertised. “The number one thing we hear from the organizations we talk to is the desire to be able to marry their vast troves of proprietary knowledge with … generative AI models — to support use cases that range from conducting research to writing to selling and beyond,” Moody and Xu told ZebethMedia in an email interview. “The magic of Mem is that we bring together your own private and proprietary data along with state-of-the-art generative language models to unlock truly personalized, factual outputs. We combine knowledge sources across the individual, team and organizational levels, leading to significantly better performance across the board.” Mem recently launched Mem It for Twitter, which allows users to save threads, get AI-generated summaries of their contents and see suggestions for similar tweets. It’s also continuing to refine Mem X, Mem’s built-in work assistant, with new features like Smart Write and Smart Edit, which leverages AI to generate text based on a prompt, summarize files, generate titles for documents and let users use natural language commands to edit or format text. Mem’s AI-powered writing tools, which are launching in preview soon. The plan for the foreseeable future is to increasingly lean into these sorts of AI-powered experiences, Moody and Xu say, with support from OpenAI through the OpenAI Startup Fund. OpenAI Startup Fund participants receive early access to new OpenAI systems and Azure resources from Microsoft in addition to capital. “OpenAI is obviously leading the wave of technological revolutions that we are riding,” Moody and Xu said. “This makes the OpenAI Startup Fund the ideal partner for what we’re building — for both the technical expertise and strategic guidance they bring to the table.” OpenAI COO Brad Lightcap, who also manages the OpenAI Startup Fund, added in an emailed statement: “Mem uses powerful AI to make knowledge workers more productive by removing the tedium and drudgery of organizing and accessing information, ultimately allowing people to focus on the parts of their work that matter. Their vision aligns squarely with our goal at the OpenAI Startup Fund to accelerate companies using AI to enhance productivity and, more broadly, human potential.” Mem competes with a number of companies seeking to tackle the same knowledge-finding and notes-organizing challenges. In enterprise search, there’s Glean, which recently raised $100 million in a venture equity round. On the knowledge management side, Atlassian’s wiki-like collaborative workspace Confluence and Notion, which was valued at $2 billion in 2020, still dominate. But Moody and Xu argue that 16-employee Mem has an advantage in that it’s “self-organizing,” ostensibly resulting in less manual curation and labor. While they declined to reveal Mem’s revenue or the names of any major customers, they assert that Mem is successful, owing to its AI-driven tech. “We’re confident in our unique approach to self-organizing and generative knowledge management. … Our personalized machine learning models not only help knowledge workers stay organized automatically, but also go beyond simply helping find things — we actually help people do their work,” Moody and Xu said. “The shift to remote work has made effective, asynchronous knowledge sharing more important than ever, and the market slowdown has caused companies to focus on efficiency. Our AI-assisted knowledge work saves people time, and the rapid improvement in large language models gives us a further tailwind.”

OpenAI leads $23.5M round in Mem, an AI-powered note-taking app • ZebethMedia

By zebethcontrolPosted on November 9, 2022Posted in AI, Enterprise, funding, Mem, note taking, OpenAI, Robotics & AI, StartupsNo Comments

Last year, OpenAI announced the OpenAI Startup Fund, a tranche through which it and its partners, including Microsoft, are investing in early-stage AI companies tackling major problems. Mum’s been the word since on which companies have received infusions from the Fund. But today, the OpenAI Startup Fund revealed that it led a $23.5 million investment in Mem, a work-focused app that taps AI to automatically organize notes. The investment values Mem at $110 million post-money and brings the startup’s total raised to $29 million. Co-founded by Kevin Moody and Dennis Xu, Mem differentiates itself from traditional note-taking apps by emphasizing “lightweight organization,” in Moody and Xu’s words. The workflow revolves around search and a chronological timeline, allowing users to attach topic tags, tag other users and add recurring reminders to notes. Mem users can capture quick notes, send links and save images from anywhere using SMS, messaging apps and the platform’s mobile client. Collaboration features let teams share, edit and comment on notes and directly attach them to shared calendars for faster reference. Mem’s search experience uses AI to search across notes, aiming to understand which notes might be most relevant in a given moment to a particular person. Moody and Xu say the platform is designed to augment knowledge workers in their typical responsibilities, like reading through pages of information, extracting the pieces relevant to a particular question and transforming the information into an answer or a report. Image Credits: Mem There’s no doubt knowledge-seeking tasks are time-consuming. According to Gartner, professionals spend 50% of their working hours searching for information and on average take 18 minutes to locate a file (albeit the veracity of metrics like these has been challenged over the years). One source estimates that document disorganization costs businesses $3,900 per employee each year in productivity losses, making Mem an attractive proposition if the tech works as advertised. “The number one thing we hear from the organizations we talk to is the desire to be able to marry their vast troves of proprietary knowledge with … generative AI models — to support use cases that range from conducting research to writing to selling and beyond,” Moody and Xu told ZebethMedia in an email interview. “The magic of Mem is that we bring together your own private and proprietary data along with state-of-the-art generative language models to unlock truly personalized, factual outputs. We combine knowledge sources across the individual, team and organizational levels, leading to significantly better performance across the board.” Mem recently launched Mem It for Twitter, which allows users to save threads, get AI-generated summaries of their contents and see suggestions for similar tweets. It’s also continuing to refine Mem X, Mem’s built-in work assistant, with new features like Smart Write and Smart Edit, which leverages AI to generate text based on a prompt, summarize files, generate titles for documents and let users use natural language commands to edit or format text. Image Credits: Mem The plan for the foreseeable future is to increasingly lean into these sorts of AI-powered experiences, Moody and Xu say, with support from OpenAI through the OpenAI Startup Fund. OpenAI Startup Fund participants receive early access to new OpenAI systems and Azure resources from Microsoft in addition to capital. “OpenAI is obviously leading the wave of technological revolutions that we are riding,” Moody and Xu said. “This makes the OpenAI Startup Fund the ideal partner for what we’re building — for both the technical expertise and strategic guidance they bring to the table.” Mem competes with a number of companies seeking to tackle the same knowledge-finding and notes-organizing challenges. In enterprise search, there’s Glean, which recently raised $100 million in a venture equity round. On the knowledge management side, Atlassian’s wiki-like collaborative workspace Confluence and Notion, which was valued at $2 billion in 2020, still dominate. But Moody and Xu argue that 16-employee Mem has an advantage in that it’s “self-organizing,” ostensibly resulting in less manual curation and labor. While they declined to reveal Mem’s revenue or the names of any major customers, they assert that Mem is successful, owing to its AI-driven tech. “We’re confident in our unique approach to self-organizing and generative knowledge management. … Our personalized machine learning models not only help knowledge workers stay organized automatically, but also go beyond simply helping find things — we actually help people do their work,” Moody and Xu said. “The shift to remote work has made effective, asynchronous knowledge sharing more important than ever, and the market slowdown has caused companies to focus on efficiency. Our AI-assisted knowledge work saves people time, and the rapid improvement in large language models gives us a further tailwind.”

Security automation startup Veriti launches out of stealth with $18.5M • ZebethMedia

By zebethcontrolPosted on November 8, 2022Posted in AI, cybersecurity, Enterprise, funding, Security, startup, StartupsNo Comments

Veriti, a platform for unifying cybersecurity infrastructure, today emerged from stealth with $18.5 million in funding, a combination of $12 million from Insight Partners and a $6.5 million round led by NFX and Amiti. According to CEO Adi Ikan, the newly announced capital is being put toward scaling Veriti’s business operations and developing its product suite. Veriti’s launch comes as VCs continue to show enthusiasm for cybersecurity startups despite the generally unfavorable funding climate. According to PitchBook data, venture capital investments in the security sector this year eclipsed $13.66 billion — up from $11.47 billion in 2020. And the global cybersecurity market is projected to be worth over $500 billion by 2030. Founded in 2021 by Ikan and Oren Koren — both ex-Check Point executives — Veriti integrates with a company’s existing security stack to evaluate risk posture by analyzing security configurations, logs, sensor telemetries and threat intelligence feeds. The platform taps AI to identify which events might be impacting business uptime and present the root cause, as well as which security policy improvements need to be taken to remediate the impacts. “Enterprise security posture is usually sub-optimal. This is due to many reasons, including tool sprawl, increased complexity, massive amounts of data and limited resources,” Koren told ZebethMedia in an email interview. “This is what inspired us to build Veriti’s platform — to address these complexities and help IT and security stay on top of this challenge.” Koren makes the case that Veriti can augment security teams’ efforts in spotting security gaps, ultimately reducing the time spent on monitoring and maintenance tasks. The growing number of security solutions in organizations can introduce complexity because each solution has its own functions and tools to learn, he argues, while the volume of alerts issued by the solutions end up creating murky visibility into the actual security posture. Koren isn’t exactly an unbiased source. But he’s not the only one who’s observed these troubling trends in enterprise security. One recent survey of over 800 IT professionals found that almost 60% were receiving over 500 cloud security alerts per day, and that the alert fatigue created by the volume caused 55% to miss critical alerts on either a daily or weekly basis. “While affording more expansive security capabilities, the proliferation of security solutions creates room for misconfigurations that can result in inadvertent security gaps and adversely impact the business by blocking legitimate applications and users,” Ikan said via email. “IT and security leadership today have a poor idea of the true utilization of security investments and of the effective security posture of their organizations.” Veriti’s challenge will be demonstrating that its approach is superior to the other security posture-analyzing platforms on the market. Rival vendor Secureframe provides a service that integrates with cloud providers and apps to understand its customers’ security postures. Hunters, another competitor, aims to automate the threat-hunting process by taking in data from networking and security tools to detect stealth attacks. It’s very early days for Veriti — Koren wouldn’t reveal the size of the company’s customer base or current revenue. But he’s betting that Veriti’s tech expertise will help it stand out from the pack. “By leveraging modern techniques like machine learning, focusing on automation, we aim to provide a way for modern teams to maximize security posture while minimizing issues that impact business uptime,” he said. As the idiom goes: time will tell.

Dataloop secures cash infusion to expand its data annotation tool set • ZebethMedia

By zebethcontrolPosted on November 3, 2022Posted in AI, data labeling, Enterprise, machine learning, Robotics & AI, startup, StartupsNo Comments

Data annotation, or the process of adding labels to images, text, audio and other forms of sample data, is typically a key step in developing AI systems. The vast majority of systems learn to make predictions by associating labels with specific data samples, like the caption “bear” with a photo of a black bear. A system trained on many labeled examples of different kinds of contracts, for example, would eventually learn to distinguish between those contracts and even extrapolate to contracts that it hasn’t seen before. The trouble is, annotation is a manual and labor-intensive process that’s historically been assigned to gig workers on platforms like Amazon Mechanical Turk. But with the soaring interest in AI — and in the data used to train that AI — an entire industry has sprung up around tools for annotation and labeling. Dataloop, one of the many startups vying for a foothold in the nascent market, today announced that it raised $33 million in a Series B round led by Nokia Growth Partners (NGP) Capital and Alpha Wave Global. Dataloop develops software and services for automating aspects of data prep, aiming to shave time off of the AI system development process. “I worked at Intel for over 13 years, and that’s where I met Dataloop’s second co-founder and CPO, Avi Yashar,” Dataloop CEO Eran Shlomo told ZebethMedia in an email interview. “Together with Avi, I left Intel and founded Dataloop. Nir [Buschi], our CBO, joined us as third co-founder, after he held executive positions [at] technology companies and [lead] business and go-to-market at venture-backed startups.” Dataloop initially focused on data annotation for computer vision and video analytics. But in recent years, the company has added new tools for text, audio, form and document data and allowed customers to integrate custom data applications developed in-house. One of the more recent additions to the Dataloop platform is data management dashboards for unstructured data. (As opposed to structured data, or data that’s arranged in a standardized format, unstructured data isn’t organized according to a common model or schema.) Each provides tools for data versioning and searching metadata, as well as a query language for querying datasets and visualizing data samples. Image Credits: Dataloop “All AI models are learned from humans through the data labeling process. The labeling process is essentially a knowledge encoding process in which a human teaches the machine the rules using positive and negative data examples,” Shlomo said. “Every AI application’s primary goal is to create the ‘data flywheel effect’ using its customer’s data: a better product leads to more users leads to more data and subsequently a better product.” Dataloop competes against heavyweights in the data annotation and labeling space, including Scale AI, which has raised over $600 million in venture capital. Labelbox is another major rival, having recently nabbed more than $110 million in a financing round led by SoftBank. Beyond the startup realm, tech giants, including Google, Amazon, Snowflake and Microsoft, offer their own data annotation services. Dataloop must be doing something right. Shlomo claims the company currently has “hundreds” of customers across retail, agriculture, robotics, autonomous vehicles and construction, although he declined to reveal revenue figures. An open question is whether Dataloop’s platform solves some of the major challenges that exist in data labeling today. Last year, a paper published out of MIT found that data labeling tends to be highly inconsistent, potentially harming the accuracy of AI systems. A growing body of academic research suggests that annotators introduce their own biases when labeling data — for example, labeling phrases in African American English (a modern dialect spoken primarily by Black Americans) as more toxic than the general American English equivalents. These biases often manifest in unfortunate ways; think moderation algorithms that are more likely to ban Black users than white users. Data labelers are also notoriously underpaid. The annotators who contributed captions to ImageNet, one of the better-known open source computer vision libraries, reportedly made a median of $2 per hour in wages. Shlomo says it’s incumbent on the companies using Dataloop’s tools to affect change — not necessarily Dataloop itself. “We see the underpayment of annotators as a market failure. Data annotation shares many qualities with software development, one of them being the impact of talent on productivity,” Shlomo said. “[As for bias,] bias in AI starts with the question that the AI developer chooses to ask and the instructions they supply to the labeling companies. We call it the ‘primary bias.’ For example, you could never identify color bias unless you ask for skin color in your labeling recipe. The primary bias issue is something the industry and regulators should address. Technology alone will not solve the issue.” To date, Dataloop, which has 60 employees, has raised $50 million in venture capital. The company plans to grow its workforce to 80 employees by the end of the year.