TechCrunch+

In 2019, Ring launched the Peephole Cam, a camera that fits over existing door peepholes to record goings on outdoors, in apartment building hallways and so on. Priced at $199, the Peephole Cam failed to catch on, leading Ring to discontinue it in 2021.

But now, Ring’s giving it another go — the Peephole Cam made a return this morning at CES 2023. While the same model, it now starts at $129 and ships with software that brings its capabilities in line with the rest of Ring’s product portfolio.

“After removing the Peephole Cam from our inventory, we heard from a number of customers who still were interested in this device and wanted to secure their front door with a Ring Doorbell but were not able to install one of Ring’s existing doorbells,” Ring CTO Josh Roth told TechCrunch in an email interview. “We are pleased to be able to offer the Ring Peephole Cam at a lower price than before to better serve our customers.”

The Peephole Cam — which Ring insists can be installed without permanent modifications to a door — offers motion detection, a doorbell button, 1080p video, two-way talk, real-time streaming video and Privacy Zones (areas users can designate to black out from their camera’s field of view). Exclusive to the Peephole Cam is an adjustable impact sensor to detect when a door is being “physically interacted with”; when the Peephole Cam senses vibrations, it’ll alert that someone’s knocking on the door and begin recording.

Predictably, the Peephole Cam works with Alexa, letting owners send announcements or sound effects to Alexa-enabled devices when a knock, motion or doorbell ring is detected. A Peephole Cam-detected knock or motion can also be set to trigger smart home routines, for example switching on connected lights and closing motorized window blinds.

When asked about the Peephole Cam’s privacy features, Roth noted that the doorbell has built-in cover slides to prevent a passerby from looking through the peephole and a toggle for audio recording. But that probably won’t allay the fears of consumer advocates who’ve argued that the company’s devices are a security threat. As TechCrunch previously reported, Ring has a history of sharing footage with the government without users’ permission, workingclosely with police departments around the U.S. and being generally reluctant to disclose its connections with law enforcement.

Ring brings back the Peephole Cam, now starting at $129 by Kyle Wiggers originally published on TechCrunch

Amazon’s conversational voice assistant, Alexa, has made its way into many electric vehicles in recent years. Now, along with ordering Alexa to ‘play Despacito’ or ‘call Mom,’ EV drivers in the U.S. with Alexa-enabled cars will be able to ask Alexa to find the nearest public charging spot.

Charging network EVgo will be the first to partner with Amazon to offer this service later in 2023. Aside from a list of available public charging stations, Alexa will help drivers in the U.S. navigate to their chosen station and initiate charging. Drivers can even pay through a linked account simply by saying, “Alexa, pay for my charge,” according to Amazon, which announced the integration at CES 2023 in Las Vegas on Thursday.

“The EV charging experience is a lot more fragmented than for gas customers, who can pretty much stop at any location,” said Anes Hodžić, vice president at Amazon’s Smart Vehicles group, in a statement. Hodžić noted that EV drivers often fumble through different apps and maps to find an available charger, taking into account factors like real-time availability, distance, remaining battery, charging speed, plug type and payment options.

Alexa’s EV charging service connects drivers with Alexa-enabled vehicles and automotive accessories like Echo Auto to over 150,000 public stations in the U.S. powered by EVgo and other operators. When a driver asks Alexa to find a station, they’ll receive a list of nearby locations with availability by plug type, as well as the time and distance to arrival. Once the driver chooses a location, Alexa can provide navigational instructions.

“At EVgo, we’re committed to making EV driving infrastructure convenient, reliable, and affordable for all types of drivers,” said Cathy Zoi, EVgo’s CEO. “Mass adoption of EVs is underway, and this collaboration between EVgo and Amazon will make charging seamless for even more EV drivers.”

At present, EVgo has more than 850 fast charging stations across the country, but it’s building out more. In June, EVgo partnered with General Motors to build more than 3,250 fast chargers in the U.S. by 2025.

Alexa, find me an EV charging station by Rebecca Bellan originally published on TechCrunch

Riding the generative AI wave, Picsart, the developer behind various photo and video editing apps for the web and mobile devices, is introducing a new iOS app that transforms photos and drawings into digital art. Called SketchAI, the app lets users sketch a picture or upload an existing image and apply different artistic styles to it.

SketchAI is easy enough to use. It features several pre-selected styles that can applied to creations, including ink drawing, pencil sketch, and the artist-inspired “Da Vinci” and “Van Gogh.” In addition to sketching or uploading a photo, users can add a prompt describing an image (e.g. “Boat at night pointing by Aivazovsky”) to enhance the generated results.

SketchAI offers five free creations. Unlocking unlimited generations requires a subscription ranging from $5.99 per week to $17.99 per month or $69.99 per year.

“We’re planning a lot in the way of making the user experience, processing and image quality better, and adding more prompts and artistic styles,” Picsart VP of product Lusine Harutyunyan told TechCrunch in an email interview. “We’re also thinking about adding features for the user to enhance and play with their results, as Picsart is known for its powerful and fun editing tools. Along with building on our main platform, we’re excited to offer more unique entry points for creative technologies like this.”

SketchAI joins Picsart’s roster of generative AI tools including AI Avatar, which makes custom AI-generated profile pics from selfies, and arrives as art-generating AI apps attract controversy both from users and the artistic community. Lensa’s recently-launched, viral avatar creator came under scrutiny for its biases toward the sexualized depiction of women. Meanwhile, on the art portal ArtStation, which earlier this year began allowing AI art for the first time, members widely protesting by placing “No AI Art” images in their portfolios — asserting that AI-generated art threatened the artistic integrity of the platform.

Harutyunyan didn’t deny that generative AI systems have their issues, even admitting that the system driving SketchAI — an open source model called Stable Diffusion — could reproduce biases in the artwork it creates. But he argued that SketchAI and generative AI as a whole will evolve and improve over time as more people use the tech and additional models become available.

“These are very early days for generative AI as a whole and this technology will continue to evolve quickly and we will continue to adhere to industry standards and best practices as it does. Our goal is to empower creators and we support artists everywhere,” Harutyunyan said.

Stable Diffusion, which is trained on images from around the web, including from art communities, has spread like wildfire in recent months. Lensa and DeviantArt use it to generate images, as does game developer Latitude and countless others. Most of the use cases are harmless enough. But some groups have wielded Stable Diffusion to create objectionable content like depictions of violence and pornographic, nonconsensual celebrity deepfakes.

Stability AI was even the subject of a recent critical letter from U.S. House Representative Anna G. Eshoo (D-CA) to the National Security Advisor (NSA) and the Office of Science and Technology Policy, in which she urged the NSA and OSTP to address the release of “unsafe AI models” that “do not moderate content made on their platforms.”

Harutyunyan says that Picsart fine-tuned and optimized Stable Diffusion for “quality, original image strength” in SketchAI and put filters in place to prevent some “unsafe” art from being generated. Picsart hasn’t, however, established a way for artists who don’t wish SketchAI users to create art in their distinctive styles to “opt out.” In recent months, artists like Hollie Mengert and Greg Rutkowski — whose names have become some of the most commonly used prompts in Stable Diffusion-powered apps — havedecried what they see as poor AI imitations that are nevertheless tied to their work.

Stability AI, the startup largely funding the development of Stable Diffusion, recently bowed to pressure, signaling that it would allow artists to opt out of the data set used to train the next-generation Stable Diffusion model. Harutyunyan says that Picsart will consider adopting the model for SketchAI once it’s released.

“The baseline model we’re using is trained on data, not specific material and not reproducing a specific artist’s work,” Harutyunyan said.

A word of warning to those who’d share their SketchAI creations publicly, though: Picsart says that it can’t guarantee users will be able to claim copyright over them. That’s because the copyright status of AI-generated artwork is somewhat in flux at the present. The U.S. Patent and Trademark Office (USPTO) recently moved to revoke copyright protection for an AI-generated comic, saying that copyrightable works require human authorship. And a class action lawsuit working its way through the U.S. court system alleges that generative AI violates intellectual property laws by regurgitating portions of the copyrighted data used to develop it.

The copyright issue has spooked platforms like Kickstarter and Getty Images, both of which in recent months have banned — either partially or fully — AI-generated art and tools to create it out of fear of the legal ramifications.

“As between Picsart and the user, the user owns the rights in the content,” Harutyunyan said. “However, users should be aware of the inherent limitations that come along with generative AI.”

Picsart’s AI-powered SketchAI app turns images and outlines into digital art by Kyle Wiggers originally published on TechCrunch

Immunefi’s Crypto Losses 2022 report found over $3.9 billion was “lost” last year. While that might seem like a whopping amount of capital to lose track of, it’s down 51.2% compared to 2021, when over $8 billion was stolen, the report found.

Crypto losses are defined as a combination of hacks and alleged fraud incidents, Adrian Hetman, tech lead of the triaging team at Immunefi, previously told TechCrunch. In 2022, the majority of losses, or $3.77 billion, were from hacks across 134 specific incidents. About $175 million was lost to fraud across 34 incidents in the same time frame.

Both decentralized finance (DeFi) and centralized finance (CeFi) experienced major catastrophic events, including the implosion of the Terra/LUNA ecosystem and the downfall of centralized crypto exchange FTX. But overall, DeFi was the main target for (successful) exploits at over 80%, Immunefi stated.

DeFi losses increased 56.2% from over $2 billion across 107 incidents in 2021 to $3.18 billion across 155 incidents in 2022. CeFi losses, meantime, fell 87.3% from $6 billion across nine incidents in 2021 to $768.8 million across 13 incidents in 2022.

The two most targeted blockchains last year were BNB Chain — crypto exchange Binance’s blockchain ecosystem — and the layer-1 blockchain Ethereum, with 65 and 49 incidents, respectively. Together, BNB Chain and Ethereum represented over half of the blockchain attacks at 63.3%. Trailing behind the two was Solana, with 12 incidents, or 6.7% of total attacks in 2022.

Looking back, every quarter had a handful of multimillion-dollar losses, some bigger than others. While each quarter had its losses, the fourth quarter saw the most, with $1.62 billion in total losses across 55 incidents, accounting for almost half of the total losses in the year.

But five major exploits, Ronin Network’s $625 million, Wormhole’s $326 million, Nomad’s $190 million, BNB Chain’s $570 million and FTX’s $650 million, accounted for about 60% of all losses in 2022.

Roughly 5%, or $204 million, of total losses were recovered in 2022.

Looking forward to 2023, it’s expected that crypto “losses” will be in the billions again as more players enter the space and capital continues to pour in. Fixing this long term will be a product of enhanced security measures, something not all projects, blockchains, protocols and other digital asset entities have prioritized.

There are also bug bounty and security services platforms that aim to protect web3 businesses and their users — but until these are implemented across the industry as a standard, more will fall to these hacks and fraudulent activities.

Crypto losses in 2022 dropped 51% year on year to $4B by Jacquelyn Melinek originally published on TechCrunch

Not everyone believes that the metaverse — social, VR-centric worlds — has staying power. But MeetKai is among the steadfast optimists. Founded in 2018, the startup initially focused on conversational AI, aiming to build cutting-edge, top-of-the-line voice recognition and speech synthesis tech. It soon broadened its focus, introducing products to help developers build components of metaverse worlds, including “intelligent” NPCs and (less excitingly) ads.

This year at the 2023 Consumer Electronics Show, MeetKai is launching several new platforms geared toward metaverse creators and users, including a way to digitize 3D spaces and buildings using any internet-connected device with a camera.

The 3D-digitizing service, called MeetKai Reality, can bring various objects and spaces from the real world into the metaverse, says MeetKai. After tapping a camera-equipped device to record a few seconds of video, users upload the footage to MeetKai Reality, which renders the captured space in VR.

While not novel — platforms like Coohom, Matterport and even Zillow could already render spaces from photos — MeetKai CEO and co-founder James Kaplan asserts that MeetKai Reality is the first photo-to-rendering solution tailored for metaverse use cases. “We want to unlock the same cost and time savings for everyone else looking to build in the metaverse,” he said in a statement, noting the potential applications in real estate, interior design, architecture, engineering and retail.

MeetKai’s second new product, MeetKai Metaverse Editor, is a bit more differentiating. It allows users without coding experience to build structures and spaces in the metaverse, optionally collaborating with others in real time as they do so. Metaverse-building tools exist, to be sure — startups like Tripolygon provide them. But not all are no-code and some remain vaporware, Kaplan currently points out.

“MeetKai’s metaverse tools can be used a la carte … We want to unlock the same cost and time savings for everyone else looking to build in the metaverse,” Kaplan said.

To round out the new product suite, MeetKai is launching MeetKai Cloud AI, which allows developers to plug a cloud-driven voice assistant into an existing virtual environment. Kaplan claims these assistants — which take the form of avatars that can converse on any number of predefined topics — have “real time reasoning capabilities” (although the jury’s out on that) and can “leverage any form of multimedia to instantly interact with end users.”

In the future, MeetKai plans to develop its own AI language system along the lines of OpenAI’s ChatGPT, which it expects will vastly improve the quality of conversations with its avatars.

“I’m very happy to see our company deliver the world’s leading metaverse and AI technology products, which set a new mark and raise the technology bar in a very competitive industry,” MeetKai co-founder and executive chairwoman Weili Dai said in a press release. “Our vision and mission for MeetKai is to deploy these capabilities to the masses by offering groundbreaking and affordable solutions, accessible from web browsers and any device — like phones, tablets, computers, large screens, not just VR — available anywhere in the world for greater impact and better lives for all.”

MeetKai is competing with hundreds of companies for dominance in the metaverse space, but it appears to be holding its own, claiming that over 50 million people actively use its tech. Recently, the startup announced a deal with the Los Angeles Chargers to develop new in-stadium and at-home experiences, including an “AI-based” locker room tour.

To date, MeetKai has raised over $20 million in venture capital (per VentureBeat) and has roughly 40 employees.

MeetKai launches new room-scanning tech and metaverse builder by Kyle Wiggers originally published on TechCrunch

Data science has long been the domain of hardcore data professionals who understand the complex frameworks and languages involved, but those professionals are in notoriously short supply.

Fortunately, the landscape of tools and frameworks is constantly evolving, and in 2023 I predict new developments that will alleviate challenges for data teams and businesses alike.

On the one hand, the long-heralded citizen data scientists will finally play a greater role in analytics thanks to sheer necessity and a simplification of the tools and platforms involved. On the other hand, data professionals will start to benefit from some of these simpler tools to accelerate their work and a push for greater standardization will help the industry as a whole.

Here are four predictions for data science in the new year and how businesses can take advantage of them.

Python use will expand beyond data professionals to citizen developers

Business people can’t afford to wait for data scientists to provide the analytics they need, so they’re taking matters into their own hands. Python has become more approachable for non-professionals with the availability of preconfigured cloud runtimes and accessible tools like NumPy for numerical data, Prophet for forecasting and H3 for geospatial data. As a result, in 2023, Python use will expand beyond data professionals and into the hands of business analysts and other less technical users.

Novice Python users should not attempt to build their own runtime environments but should opt for any of the modern cloud platforms that provide built-in security and governance. Anaconda offers a popular Python distribution that helps ensure updates and dependencies are managed properly, and Snowflake installs these packages in our cloud-based Python runtime.

There are numerous online resources for non-professionals to get started with Python, including this comprehensive beginner’s guide from RealPython.

4 data science predictions for 2023 by Annie Saunders originally published on TechCrunch

This year, the Consumer Electronics Show (CES) — the celebration of all things tech and then some — brought what it usually brings to the Vegas Strip: AI-powered gadgets. Of course, AI comes in many forms, and not all that’s advertised as AI is in fact true AI. But there’s always diamonds in the rough, like an AI-equipped oven from Samsung that automatically adjusts cooking settings for the perfect bake.

As my colleagues on the ground traverse the show floor (I’m covering CES remotely this year), they’re highlighting the most interesting AI-powered tech they come across. I’ve compiled each into a handy list, which is far from definitive, but which aims to give an idea of the top AI trends this year at CES.

AI-powered oven

Samsung kicked off CES 2023 with a new oven and fridge in its Bespoke Home smart appliances lineup.

Not just any oven, the Bespoke AI Oven features an internal camera and sensors that proactively recognize more than 80 dishes and ingredients to optimize cooking settings. It can send a warning notification if it detects food might be burning. And if you’re stumped on what to cook, the Bespoke AI Oven recommends meals based on ingredients at home.

The Bespoke AI Oven’s camera, by the way, can stream a live feed of what’s cooking in the oven to streaming platforms like Twitch. It’s not an AI feature exactly, but eye-catching all the same.

As for the 4-Door Flex, Samsung’s newest smart fridge, it packs an AI-powered camera that scans food labels to keep track of what’s in stock — and not in stock. Samsung’s previous-gen fridge had the same (as do rival models from LG), but the novelty of this year’s 4-Door Flex is a larger touchscreen on one of the doors.

AI-powered TVs

Not to be outdone, Samsung also introduced a new lineup of TVs with an AI-powered “auto-HDR” mode.

The company’s Neo QLEDs offer Auto HDR Remastering, which Samsung says “uses AI deep learning technology to analyze and apply real-time high dynamic range (HDR) effects on standard dynamic range content on a scene-by-scene basis.” (HDR increases the contrast between the brightest whites and darkest blacks a display can produce.) The system compares non-HDR content with HDR material and uses machine learning to analyze the differences and then make the conversion.

Rival LG debuted TVs with AI, too. The 2023 Z3, G3, and C3 OLED TVs feature a mode that can automatically detect and refine certain objects and like faces, to have more “lifelike,” colorful qualities. Another mode, AI Sound Pro, attempts to emulate a surround sound setup using only the built-in stereo TV speakers.

Upscaling GPUs

Nvidia’s using AI to upscale old, blurry YouTube videos.

During its CES presser, the company demoed RTX Video Super Resolution, a feature coming next month to all Nvidia 30- and 40-series GPUs. When enabled, It’ll upscale videos played in Chrome and Edge on a PC to the equivalent of 4K, similar to the AI upscaling feature that’s long been available on the Shield TV and Shield TV Pro.

According to PCGamer (via The Verge), RTX Video Super Resolution — which also sharpens and reduces artifacts in videos — will support any video with a resolution of between 360p and 1440p up to a frame rate of 144Hz.

Fatigue-battling watch

Wearable brand Citizen dipped its toes into the AI space this year with its second-gen CZ Smart watch.

The 2023 CZ Smart — which runs Wear OS 3, the newest release of Google’s firmware for wearables — sports what Citizen’s calling a “self-care advisor” driven by AI. Leveraging tech from IBM Watson (IBM’s AI research division), the CZ Smart attempts to determine a wearer’s “chronotype” after collecting roughly a week’s worth of sleep data to provide tips that help to combat fatigue.

To improve the relevancy of the recommendations, users can take gamified tests based on NASA’s Psychomotor Vigilance Test, which was originally developed to give astronauts on the International Space Station feedback on their fatigue levels. The CZ Smart tabulates the results to calculate an “alert score” capturing the wearer’s level of tiredness at a particular moment.

It’s worth noting that the CZ Smart isn’t the first wearable to track sleep and give suggestions to improve it — far from it. Samsung, Apple and Fitbit among others offer features on their various wearables to analyze and track sleep. But Citizen’s making the case that the CZ Smart is more holistic than most.

Flashy lighting

For gamers in pursuit of a more immersive experience, Govee introduced the AI Gaming Sync Box Kit at CES 2023.

It’s a bit of a mouthful, but the tech behind the AI Gaming Sync Box Kit — Govee’s answer to the Philips Hue Play HDMI Sync Box and Nanoleaf’s recently launched Sync+ — sounds promising. Using AI trained on footage of “dozens” of different video games, the set-top box syncs (via HDMI and Wi-Fi) Govee-branded smart lights to whatever games you’re playing. One of 30 or so lighting effects trigger for certain in-game achievements and on-screen events.

As the company explains in a press release: “For example, when accomplishing a ‘triple kill’ in a competitive shooting game, a customized corresponding in-game lighting effects will show at once with near-zero delay. With Govee AI, products don’t transfer data to the cloud but instead directly transmit it back locally in real time, ensuring the utmost protection of user privacy.”

Govee says that its AI-based tech mitigates the issues typically experienced by lighting sync boxes, like “low color-capture accuracy” and slow performance. The jury’s out on that. But in any case, it’s certainly one of the more innovative approaches we’ve seen.

Bird-friendly feeder

Bird Buddy, the startup my colleague Devin Coldewey covered a number of years ago, is back with a new and improved model of its “smart” feeder that snaps photos of bird visitors and beams them to a companion app.

Equipped with a high-speed camera, the next-gen Smart Hummingbird Feeder can take photos and videos of birds with wing speeds of up to 60 mph. The camera’s triggered by motion and sends photos through an AI system trained on millions of photos to identify around 1,000 species of birds.

Future owners of the Smart Hummingbird Feeder — it’s a prototype currently, and hasn’t yet been priced — will also benefit from what Bird Buddy’s calling Heartbeat Map, which lets people track bird sightings in real time. Meanwhile, the aforementioned Bird Buddy companion app allows owners to build out their collection of identified birds, track birds’ visits over time, learn about their habits and share photos with the community.

Self-driving stroller

You’ve heard of self-driving cars, but what about self-driving strollers?

The Ella smart stroller by Canadian startup Glüxkind is a “hands-free” motorized buggy that uses AI to detect obstacles and danger on the pavement. Packed with sensors including cameras, the Ella can drive itself when empty, keep up with parents while they hold their infants and stop automatically if it encounters an obstacle.

The buggy’s cameras track moving and stationary objects such as people and bikes, alerting parents to potential collision dangers through sounds and flashing lights in the handle. As for the Ella’s motors, they aid when manually pushing the stroller, assisting with inclines and automatically braking when going downhill or stationary.

The Ella isn’t lacking in creature comforts, either, featuring an automatic “rock-my-baby” feature that sways the carriage back and forth on the stroller’s wheels and a built-in, customizable white noise machine. There’s even a companion phone app that tracks the Ella’s routes and distance and can locate the stroller if lost.

Pepper-picking robot

Gather ’round, pepper growers. There’s a new robot in town that promises to make your life a bit easier.

Developed by Japanese agritech startup Agrist, the robot can identify and pick “harvest-ready” bell peppers even when those peppers are covered by a thick layer of leaves. Using a combination of cameras and computer vision models trained for different parts of the pepper, the robot can identify each pepper’s position, size and maturity as well as its clipping point.

The robot moves along suspension wires that can be retrofitted to existing farms. A mechanical arm does the picking and clipping, placing harvested peppers into a collection box. Agrist claims that, over time, the robot’s software can begin to predict crop volume and offer suggestions on how to optimize growth by collecting data and analyzing the number of days required for peppers to mature.

Here’s a roundup of the top AI-powered products we saw at CES 2023 by Kyle Wiggers originally published on TechCrunch

Apartment living can be a lonely life, but OneRoof is here to change that. The company offers a hyperlocal social network app meant to connect people in the same residential building around common interests like hobbies and to even organize again a common goal.

The pre-revenue company has grown up a lot since we profiled it in 2021 after OneRoof secured $1.25 million in funding. At the time, CEO Selin Sonmez told TechCrunch that the app was live in 400 buildings in New York City, and today that has increased to over 40,000 active neighbors across 1,300 buildings in New York and Miami. OneRoof has also done a soft launch in Dallas, she added.

If you are trying to picture how OneRoof works, Sonmez says the company is often compared to NextDoor. It helps people who live in the same apartment building to connect, network, exchange info, buy and sell items, help with pets, meet in real life and create a sense of community.

The free app relies on what Sonmez called “a community of champions” known as Super Neighbors — essentially the people in your apartment building who “act as a glue to their community,” she said. From that position, they are able to gain access to neighbor event sponsorships, join the exclusive Super Neighbor Club in their city and share feedback with OneRoof for future app development.

“‘Super Neighbors’ are a critical, yet neglected social circle,” she added. “Nobody knows who their neighbors are, so we are building an essential social tissue in society. Our long-term mission is to unify and empower neighbors under OneRoof and to create more resilient urban cities, both socially and economically.”

In 2021, OneRoof was just a chat room app where neighbors could talk to each other digitally. Since then it has morphed to include an in-person events feature, improved user profiles to highlight interests, a neighbor directory and a “grow” feature to help people create their own communities.

The concept seems to be appealing to apartment dwellers: In Super Neighbor buildings, OneRoof saw over 78% of users continue using the app after 12 weeks, and over 65% continue using after 24 weeks, Sonmez said. In addition, three times the in-person events, including happy hours and meeting someone for a workout, were created on OneRoof in the fourth quarter compared to the previous quarter.

The company has continued to grow, receiving between 30 and 40 Super Neighbor requests per day, and observing a 20% month over month increase in users. To be able to onboard them, Sonmez and co-founder Nikos Georgantas needed to grow their team so they made the decision to go after a new round of funding.

This was during the spring and summer, however, when the venture capital market was slowing down. As a result, Sonmez said she had to manage not only some difficult decisions, for example, slashing the marketing budget by 90% to keep OneRoof’s existing team in place. She had close to 80 investor meetings during this time and even had to field some unexpected questions.

“We were suddenly being asked questions around revenue which wouldn’t be the case for social startups our stage before the markets shifted, so it took us time to find the right partners,” she added. “There was also market ambiguity around pricing, so we didn’t price the round until there’s more clarity.”

Eventually, Sonmez found good investment partners and closed on $3.2 million in seed funding to give the company $4.45 million in total funding. Chamaeleon led the round and was joined by Dream Machine, Gaingels, General Catalyst, The Helm and a group of angel investors, including Margo Georgiadis at Synetro Group, former Andreessen Horowitz investor D’Arcy Coolican, Google Maps co-founder Lars Rasmussen and Elomida Visviki.

The funding enabled OneRoof to add four new employees and is currently looking to fill two more roles. Sonmez also plans to invest in app development, software, sponsorships and marketing.

Regarding the company’s current big milestones around engagement and retention, OneRoof has gained a lot of ground, but Sonmez still has much she wants to accomplish.

“We want to do more,” she added. “There’s hundreds of millions of people living in this country, and we want to make sure that we can grow our products and put them in the hands of anyone who has a neighbor who thinks, ‘Oh, I wish I knew who they were.’”

OneRoof grabs funding to help apartment dwellers cozy up to their neighbors by Christine Hall originally published on TechCrunch

CircleCI, a company whose development products are popular with software engineers, has urged users to rotate their secrets following a breach of the company’s systems.

The San Francisco-headquartered DevOps company said in anadvisory published late Wednesday it is currently investigating the security incident — its most recent in recent years.

“We wanted to make you aware that we are currently investigating a security incident, and that our investigation is ongoing,” CircleCI CTO Rob Zuber. “At this point, we are confident that there are no unauthorized actors active in our systems; however, out of an abundance of caution, we want to ensure that all customers take certain preventative measures to protect your data as well.”

CircleCI, which claims its technology is used by more than a million software engineers, is advising users to rotate “any and all secrets” stored in CircleCI, including those stored in project environment variables or in contexts. Secrets are passwords or private keys that are used to connect and authenticate servers together.

For projects using API tokens, CircleCI said it has invalidated these tokens and users will be required to replace them.

CircleCI, which in 2021 announced a $100M Series F at a $1.7B valuation, hasn’t shared any more information about the nature of the incident and has yet to respond to TechCrunch’s questions.

However, the company is also advising users to audit their internal logs for unauthorized access occurring between December 21, 2022 and January 4, 2023, which suggests the company’s breach began some two weeks earlier. The company on December 21 also announced that it had released reliability updates to the service to resolve underlying “systemic issues.

In 2019, CircleCI was hit by a data breach after a third-party vendor was compromised. This saw hackers compromise user data including usernames and email addresses, usernames and email addresses associated with GitHub and Bitbucket, along with user IP addresses.

In November, CircleCI said that it had also witnessed an increasing number of phishing attempts whereby unauthorized actors were impersonating CircleCI to gain access to users’ code repositories on GitHub.

CircleCI warns customers to rotate ‘any and all secrets’ after hack by Carly Page originally published on TechCrunch

At a time when there is a lot of debate going on around ChatGPT and generative AI potentially eating up jobs, Apple has launched AI-powered audio narration for select titles on Apple Books. On its website for authors, the company says that this feature will help independent authors who might not be able to convert their titles to audiobooks because of “the cost and complexity of production.”

Currently, only select titles are narrated by Apple’s own AI-generated voices. Users will see “Apple Books” in the narrator section for these titles.

While Apple says the program is for independent authors, currently they have to sign up with partner publishing companies — Draft2Digital or Ingram CoreSource — to get their book narrated by Apple’s AI voices.

Apple is already accepting submissions under the romance and fiction genres — with support for only literary, historical, and women’s fiction at the moment — through the partners mentioned above. Plus, it is starting its AI-powered voice narration work for nonfiction and self-development genres. Currently, Apple offers four voices under soprano and baritone categories: Madison and Jackson (romance and fiction); Helena and Mitchell (self-development and non-fiction). The company said that these voices are trained in specific genres, but Apple didn’t specify what training data it is using to tune them.

“Apple Books digital narration brings together advanced speech synthesis technology with important work by teams of linguists, quality control specialists, and audio engineers to produce high-quality audiobooks from an ebook file. Apple Books has long been on the forefront of innovative speech technology and has now adapted it for long-form reading, working alongside publishers, authors, and narrators,” Apple said on its website.

If you look at audiobook titles on the Apple Book store, they are often narrated by the author with other guests or voice artists pitching in at times. Apple’s AI-generation suit probably aims to remove any need for sitting down and recording the narration in a studio. But even with this process, results are not instant. The Cupertino-based company said that once an author submits a request, it takes “one to two months to process the book and conduct quality checks.”

Apple specified that if there is not enough time for post-production checks, the title is published as soon as the processing is completed. However, the company didn’t provide any details on what might be the shortcomings of such a rushed process and how the final might sound.

It’s not clear if authors are allowed to port their Apple AI-powered audiobooks to other platforms. We have asked Apple for a comment, and we will update the story if we hear back.

A report from The Guardian noted that Apple wanted to release this feature last November. But it got delayed because of industry-wide layoffs and the Elon Musk-Twitter drama.

Most platforms — apart from Audible — allow books that are narrated by AI-generated voices. However, Apple’s entry into the market might attract a lot of attention to platforms such as Matrix-backed Murf that creators to make AI-powered audiobooks. At the moment, there’s no detail about how much Apple is charging publishers (or authors) for the whole process of converting a book into an audiobook.

As Apple is getting pushback from regulators about App Store fees, and it might have to allow alternative app stores and third-party payment options that might have an impact on its revenue. So when generative AI is having its moment, Apple is taking the first steps into making it a viable cash-generating option.

Apple launches AI-powered book narrations by Ivan Mehta originally published on TechCrunch