Bird founder Travis VanderZanden officially leaves the nest

Travis VanderZanden’s slow-motion departure from Bird is now complete. The scooter rental company announced in a late-Friday news dump that the executive has stepped down from his role as chairperson of Bird’s board, “effective immediately.”

Replacing VanderZanden is John Bitove, who played a role in saving Bird’s bacon this past December via its merger with Bird Canada.

VanderZanden had led the micromobility company from its inception as its president and founding CEO, but that all changed last year when Bird’s declining stock price culminated in a delisting warning from the New York Stock Exchange. Soon after, VanderZanden stepped down from his role as president, handing over the title to Bird’s then-chief operating officer Shane Torchiana. Torchiana went on to assume the CEO post as well several months later. At the time, VanderZanden called the reorg a “long-planned transition.”

According to Bird, VanderZanden “decided to step down [from the board] to pursue other ventures.” In a similarly vague yet intriguing statement, VanderZanden added that he intends to return to his “entrepreneurial roots and incubate some new ideas.”

TechCrunch has reached out for more information on the founder’s departure and will update this story when we hear back.

Bird founder Travis VanderZanden officially leaves the nest by Harri Weber originally published on TechCrunch

from https://ift.tt/NuD8LqU
via Technews

Read More

Are corporations too influential?

Welcome to Startups Weekly. Sign up here to get it in your inbox every Friday.

This week, I’ve been doing a lot of thinking about how some of the biggest companies in the world have as much — if not more — power than entire countries. Most countries, at least, have some level of democratic oversight, but that isn’t true in the same way for companies. My question, then: In a world where the policies of, say, Facebook, YouTube and Twitter become de facto standards all around the world, should we have a greater degree of say (TC+) in what those policies are?

The other thing that’s kept me busy this week is fundraising. Alex talked with 11 VCs (TC+) about how hard it was for their companies to raise so far this year. Meanwhile, I talked with a number of founders who were really struggling to raise money. The truth is, the founders struggling the most have three things in common (TC+).

Now let’s take a look at what happened in the world of startups this week.

Notes from the security frontlines

Image Credits: Bryce Durbin / TechCrunch

The most popular story on TechCrunch in the past week was one of my own, which came with a curious backstory: Flipper Devices was founded in Moscow, Russia, in 2020, by a Ukrainian founder and a largely Russian team. I ran the headline that a “Russian hacking device” had made $80 million worth of sales, only for a bunch of PR people to get very upset with me for calling the company, which was founded in Russia and whose team is still 90% Russian, Russian. Don’t get me wrong, I get why a company making a hacking device might not want to be associated with Russia — and the company has gone to great lengths to scrub any traces of that connection from the internet. The whole story was pretty weird, and concluded with me getting an unsolicited scan of the founder’s (Ukrainian) password in my email inbox. Very curious indeed.

That sounds secure…: In a beacon of “here’s what not to do,” Lorenzo reports that an Illinois high school accidentally changed every student’s password to ‘Ch@ngeme!’. The problem? For a moment there, every student knew every other student’s password. D’oh.

Stupid and pointless: Prosecutors called for the British hacker who was responsible for the 2020 Twitter breach to serve at least seven years. Zack reports that the hacker was sentenced to 5 years behind bars. The convicted hacker described his crimes as “stupid and pointless.” Who am I to disagree?

Watching the watchers: Zack reports that Polish-developed stalkerware LetMeSpy, a phone-tracking app, says it was hacked. The leaked data included years of victims’ call logs and text messages dating back to 2013.

News you can touch. Yep, it’s hardware.

Image Credits: Getty Images

A ton of interesting things happened in startup hardware land this week. Uplift Labs signed an interesting deal with Major League Baseball to use the startup’s 3D motion tracking tech to help scout for promising players.

Fast on the heels of its previous $14 million fundraise, Realtime Robotics raised another $10 million or so, representing the third close on what now seems like a never-ending Series A financing for the manufacturing automation startup.

Apropos robotics, Brian also had a fascinating story today on how robots are learning from watching YouTube videos. If my YouTube recommendations are anything to go by, every robot in the world will very soon be expert woodworkers and do very stupid things with explosives.

Who’s a good bot? That’s right, you’re a good bot: In a, “Geez, I feel safer already” type moment, Brian reports that the House GOP discussed the use of robot dogs to patrol U.S. borders.

It flies and it counts. That’s just what it does: Kate reports that B Garage raised $20 million for its warehouse inventory drones. And as we’re talking about flying inventory drones, Brian reported that Gather AI bought drone inventory competitor Ware.

Walking? Feh, check the webcam: The lazy among us may have pointed a webcam at the oven to keep an eye on a pizza, but Devin reports that Lilz takes the same concept to a whole ‘nother level, bringing its gauge-watching smart cameras to the U.S. and raising $4 million.

Startups that are going places

Image Credits: Joby Aviation

Raise your hand if you saw this one coming (while I sit on my hands, because I really did not) — but it seems like the Tesla charging standard is gaining a foothold very quickly. First, Texas said that state-funded EV chargers had to include Tesla plugs (now known as the North American Charging Standards, or NACS), and it seems like Washington state may be following suit.

Wheeee: You couldn’t force me on board one of these things with a gun, but Joby Aviation has reasons to celebrate, as Rebecca reports that the company received a permit to fly its first eVTOL built on a production line.

Pulling the e-brake: Kate reports that Singapore’s ride-hailing firm Grab lays off over 1,100 employees, representing around 11% of its staff — its first big round of layoffs since 2020.

End of the road for Lordstown: It’s been an uphill battle for Lordstown Motors. Rebecca reports that the company is suing Foxconn, claiming fraudulent conduct that “destroyed” the American company’s business. Over on TC+, Alex ponders that there’s not a lot of SPAC deals left that didn’t come crashing down painfully and spectacularly. Canoo, anyone?

Despite all its rage, it is still just a car in a cage: Even as Lordstown implodes and a lot of the other EV companies are struggling, Faraday Future raises $90 million to keep itself alive.

Top reads on TechCrunch

Image Credits: Forcite

Foo-wee, it’s been a lively week. My personal favorite was Tim’s story about Forcite launching a $1,100 smart helmet, finally bringing a version of the decade-old Skully dream to fruition.

U so basic: Netflix decided that it had enough of letting its users skate by on the cheap, and Ivan reported that the streaming giant quietly axed its basic plan in Canada.

We totally have lots of users, promise! Some strange dodginess this week — Amanda reported that Unicorn social app IRL is to shut down after admitting 95% of its users were fake.

Yeah, saw that one coming: In my very personal opinion, Shein — and other, similar purveyors of essentially disposable clothing — is the literal worst for the environment. It seems like the company got a sheen of comeuppance, as Amanda reports that an influencer’s highly curated trip to a Chinese factory backfired.

The crowd is going Vilnius: Europe keeps investing huge sums of money into tech ecosystems, and Paul reports that Lithuania’s capital Vilnius is about to invest more than $100 million into “Europe’s largest tech campus.”

---

Get your TechCrunch fix IRL. Join us at Disrupt 2023 in San Francisco this September to immerse yourself in all things startup. From headline interviews to intimate roundtables to a jam-packed startup expo floor, there’s something for everyone at Disrupt. Save up to $600 when you buy your pass now through August 11, and save 15% on top of that with promo code STARTUPS. Learn more.

Are corporations too influential? by Haje Jan Kamps originally published on TechCrunch

from https://ift.tt/qxCdj1F
via Technews

Read More

Police want robotaxi video footage to help solve crimes

Police are starting to turn to robotaxis — specifically all of that footage captured by cameras — for video evidence to help solve crimes. While it might not be a trend quite yet, evidence suggests that the robotaxi is the new proving ground for privacy advocates and law enforcement, especially as companies like Cruise and Waymo scale to new cities.

Self-driving cars can have more than a dozen cameras capturing 360-degree views and reams of data as they navigate city streets. And it turns out, that’s attractive to government agencies looking for evidence.

For instance, Bloomberg reported Thursday that it found nine search warrants that had been issued for autonomous vehicle company Waymo’s footage in San Francisco and Arizona’s Maricopa County. Waymo is also testing in Los Angeles. Cruise, a Waymo rival that has operations in San Francisco, Phoenix, Austin and Houston, also received a warrant, according to Bloomberg.

The instances all seem reasonable — police wanted help learning more about crimes ranging from murders to robberies to an attempted kidnapping.

“Autonomous vehicles are recording their surroundings continuously and have the potential to help with investigative leads,” reads a San Francisco Police department training document, which was obtained by Vice in 2022. “Investigations has already done this several times.”

Matthew Guariglia, senior policy analyst at nonprofit digital rights group Electronic Frontier Foundation (EFF), says the issue with police being able to tap data from otherwise unassuming vehicles is that there’s not always transparency on how the data is collected and stored, and how the police can access it.

“If an autonomous vehicle rolls up to a street corner and parks for a while, how would anyone know — or not know, for that matter — if there were police standing over a Cruise operator’s desk, saying, ‘Move a little bit closer to that corner because we want footage of a drug deal,'” said Guariglia.

Cruise and Waymo both told TechCrunch that, relative to the number of miles they drive, police requests don’t happen often. When they do, the companies say they only provide police with data when there’s a warrant or a subpoena.

“We carefully review each request to make sure it satisfies applicable laws and has a valid legal process,” a Waymo spokesperson told TechCrunch. “Waymo will analyze the requested data or information to make sure it falls within the scope of the warrant. If a request is overly broad (asks for too much information), we try to narrow it, and in some cases we object to producing any information at all.”

Both companies also say they tailor the data provided to the specific subject of the warrant. For example, if a warrant asks for information to identify another vehicle, Cruise may only provide stills from a video. If a request requires video, then Cruise might provide a short clip from a single camera.

Waymo says it blurs license plates and faces of people in order to protect the privacy of bystanders who may appear in the imagery requested in the warrant. A spokesperson told TechCrunch that there are exceptions to this rule. The company might, say, unblur the license plate of an offending vehicle in the case of a hit and run, but the warrant would need to provide a detailed description of the vehicle.

Cruise did not respond in time to confirm if it has a similar policy.

“Privacy is extremely important to us which is why we disclose relevant data only in response to legal processes or exigent circumstances, where we can help a person who is in imminent danger,” Navideh Forghani, a Cruise spokesperson, told TechCrunch.

Forghani went on to say that Cruise may share information without formal processes under genuine emergency circumstances, such as amber alerts, medical emergencies or active crimes — like sexual assault, assault with a deadly weapon, robberies, active shooter events and acts of terrorism.

But Guariglia says it’s a slippery slope. In recent years, Amazon’s Ring, a doorbell and home security company, cozied up to law enforcement around the country, giving police easy access to data from its network of individual consumer products. Guariglia says a number of companies feel compelled to build tools that allow police to access their data, even though the police aren’t their customer.

“We have to ask ourselves constantly, what do the companies get out of this?” Guariglia told TechCrunch. “That exact thing might not happen with Cruise [and Waymo], but the concern is that cities offer permits for these companies to operate. What happens if cities start to look more favorably on companies that have cozy relationships with police departments?”

I’m not doing anything wrong, who cares if the police have data?

For those who say it doesn’t matter if police have access to footage because they aren’t doing anything wrong, Guariglia says, “you have no idea what you’re doing wrong.”

“People in a lot of states where it was legal to get an abortion a few months ago suddenly have to live in fear that any day now, these states could retroactively prosecute people,” he said. “And then you start to wonder about all those months where you traveled to your doctor or mental health specialist, how much data had been collected and what can law enforcement learn about me when I didn’t think I had anything to hide?”

Self-driving cars not only operate in cities. Autonomous trucks are also being tested on highways in hopes of eventually launching commercial operations with a safety driver behind the wheel. In the U.S., where there are now 14 states that have completely banned abortion, and Idaho has restricted travel out of state for abortions, there is a real fear that local law enforcement could attempt to use AV footage to prosecute people who seek reproductive freedom.

Abuse of power is a main concern, and one with precedent.

In 2020, the Electronic Frontier Foundation sued the SFPD for conducting mass surveillance of Black Lives Matter protesters using a downtown business district’s camera network. The records obtained by the EFF showed that SFPD received real-time live access to hundreds of cameras and “data dump” of camera footage amid demonstrations against police violence.

Going back further, privacy advocates point to the revelations revealed by Edward Snowden in 2013, a contractor for the National Security agency who leaked information on the PRISM program and bulk collection of phone metadata, which collected data on millions of people without targeted suspicion of wrongdoing.

And of course, any time you increase police access to surveillance and put communities under a microscope, the people who will get hurt are already marginalized communities, says Guariglia.

Tech has tested the boundaries of surveillance before

Aside from Ring, police departments in the U.S. also use automatic license plate readers to track the movements of vehicles and can use geofence warrants issued by a court to search databases and find all active mobile devices within a particular geofenced area. And, of course, police can request footage from the millions of security cameras in businesses and residences around the country.

Privacy advocates say that adding rolling networks of autonomous vehicle cameras and data to that cocktail of surveillance is cause for concern. At a minimum, there’s the potential of violation of basic rights to privacy. But the use of video footage for surveillance also opens the door for the abuse of power, such as scope creep, or surveillance of individuals unrelated to the crime being investigated. It can also lead to a chilling effect, wherein people might alter their behavior or expressions of free speech if they fear they are constantly being monitored.

Police want robotaxi video footage to help solve crimes by Rebecca Bellan originally published on TechCrunch

from https://ift.tt/mkWOUK8
via Technews

Read More

TechCrunch roundup: Vertical AI crypto losses decline machine learning investor survey

According to Paris Heymann, a partner at Index Ventures, the current wave of AI-powered products and services can be sorted into three layers:

• Foundational models
• AI infrastructure
• AI applications

“Some of these applications will be broadly horizontal,” he writes in TC+, “but many AI applications will also be vertical, or industry-focused.”

---

Full TechCrunch+ articles are only available to members
Use discount code TCPLUSROUNDUP to save 20% off a one- or two-year subscription

---

In this market map, Heymann looks at AI stack startups exploring myriad aspects of this explosive sector, shares thoughts on where “Vertical AI” is heading, and offers advice on how SaaS startups should approach embedding features and functionality.

“Proprietary data and distribution will be a winning combination in the race to build both horizontal and vertical AI applications,” he predicts.

We’re publishing on a lighter schedule to commemorate Independence Day, so I’ll return on Friday, July 7 with a new TC+ roundup.

Have a great week!

Walter Thompson
Editorial Manager, TechCrunch+
@yourprotagonist

Vertical AI: The next logical iteration of vertical SaaS

How confidential computing could secure generative AI adoption

Image Credits: MirageC (opens in a new window) / Getty Images

The generative AI land rush has created a new challenge: how can enterprises use proprietary data to build powerful models while maintaining security and privacy?

“Protecting training data and models must be the top priority,” says Anjuna CEO and co-founder Ayal Yogev.

“It’s no longer sufficient to encrypt fields in databases or rows on a form.”

How confidential computing could secure generative AI adoption

Big tech corporate venture capital generative AI startups

Image Credits: Bryce Durbin / TechCrunch

Generative AI startup Typeface emerged from stealth this year, but the $100M Series B round that just boosted it to a billion-dollar valuation signals that “big tech companies are busy wielding their most powerful weapons: checkbooks,” writes Alex Wilhelm.

In The Exchange, he examines how companies like Microsoft, Salesforce and other corporate venture entities are buying their way into AI-adjacent products and services.

“So long as your founding team has an AI pedigree and a plan to sell lots of AI tech to big companies, you can expect a flood of cash from big tech funds and traditional venture investors alike.”

Big Tech corporate venture capital generative AI startups

Crypto losses halved in Q2 2023 to $204M

Image Credits: Getty Images

Are the rug pulls, exploits and pump-and-dump schemes that have scammed so many crypto consumers becoming a thing of the past?

According to a report by De.Fi that used information from REKT, “losses in the second quarter were 55% narrower than in Q1 2023, when the industry lost a whopping $462.3 million to hacks and scams,” writes Jacquelyn Melinek.

Crypto losses halved in Q2 2023 to $204M

4 VCs illustrate why there’s good reason to be optimistic about the machine learning startup market

Image Credits: Bryce Durbin / TechCrunch

Training the expansive models that power AI applications is thirsty work: machine learning startups can charge clients for anything from optimizing ad campaigns to generating gargantuan data sets.

Kyle Wiggers surveyed four investors to get their thoughts on whether “the hype cycle in ML dying down or going strong,” discuss technical roadblocks holding the industry back, and get their advice for founders who are building in the sector:

• Lonne Jaffe, managing director, Insight Partners
• Jerry Chen, partner, Greylock
• Ashish Kakran, principal, Thomvest
• Janelle Teng, VP, Bessemer Venture Partners

4 VCs illustrate why there’s good reason to be optimistic about the machine learning startup market

Ask Sophie: How do we relocate Ukrainian and Russian team members to the US?

Image Credits: Bryce Durbin/TechCrunch

Dear Sophie,

Our startup employs about 30 people globally through a combination of direct and co-employment based on their country.

Over the last year and a half or so, we helped several team members relocate from Ukraine and Russia to various non-Schengen countries such as Georgia, Taiwan, Thailand, Turkey, and Uzbekistan.

We realize it’s more expensive if we bring these employees to the U.S., but our startup will be more successful. How do we bring them here?

— Meaningful Moneymaking

Ask Sophie: How do we relocate Ukrainian and Russian team members to the US?

Fund of funds are starting to play a different role for venture LPs

While traditional fund of funds have fallen out of favor, there is still LP appetite for innovative approaches.

After reviewing the last few years of declining fund of funds activity, Rebecca Szkutak found that several factors are leading limited partners to look elsewhere.

“There has been way more opportunity to invest in a VC than there has ever been in the past,” said Kyle Stanford, a senior venture analyst at PitchBook.

“For new LPs coming into the market, they didn’t need to go to a fund of funds to get access.”

Fund of funds are starting to play a different role for venture LPs

Pitch Deck Teardown: Super.com’s $60M Series C deck

Image Credits: Super.com (opens in a new window)

Once known as Snapcommerce, travel/fintech/e-commerce startup Super.com has acquired almost $200M in funding, including an $85M Series C in April 2023.

The company, which offers a credit-building cashback card, shared the 17-slide pitch deck it used to close its latest round with TC+:

• Cover slide
• Mission slide
• Traction slide 1
• Traction slide 2
• Team slide
• Customers (interstitial slide)
• Customer profile
• Target customers / market size
• A Savings Super App (interstitial slide)
• Why Now slide
• “Why a super app” slide
• Product slide
• Product road map slide
• User behavior slide
• Solution slide (“SuperCash is core to the experience”)
• Value Proposition slide (“Personalized Experiences”)
• The Ask slide

Pitch Deck Teardown: Super.com’s $60M Series C deck

11 VCs reveal how hard it was for their startups to fundraise in H1 2023

Image Credits: tommy (opens in a new window) / Getty Images

Alex Wilhelm surveyed 11 investors to learn about the headwinds their portfolio companies leaned into in the first half of 2023.

“From their answers, it appears a startup’s ability to fundraise in today’s climate is based on several key factors, including capital efficiency, the market and its needs,” he writes.

11 VCs reveal how hard it was for their startups to fundraise in H1 2023

TechCrunch+ roundup: Vertical AI, crypto losses decline, machine learning investor survey by Walter Thompson originally published on TechCrunch

from https://ift.tt/XcI1TAY
via Technews

Read More

Say goodbye to Q2 and the crypto hacks scams and rug pulls that came with it

Follow me on Twitter @Jacqmelinek for breaking crypto news, memes and more.

Welcome back to Chain Reaction.

As if the pessimism around crypto weren’t enough, the industry is facing yet another quarter of hackers and scammers looking to make a quick buck. And to make things worse, it’s getting harder to trace and recover lost funds as well.

According to a new report, only $4.9 million was recovered of the $204.3 million the industry lost to hacks, scams and rug pulls in the second quarter.

The report, by web3 “super app” and antivirus solution De.Fi and data from REKT database, detailed that so far this year, the industry had recovered about $183 million, or nearly 28% of the $666.5 million lost to scams and hacks.

The report also found that exploits and rug pulls accounted for $55.3 million and $47.3 million, respectively, in Q2, highlighting that risks through bad actors are “rampant in equal measure.”

The TLDR? Be careful out there, because hackers are still hackin’ and scammers are still scammin’ — even in a bear market.

This week in web3

Q2 failed to bring a funding reprieve for web3 startups and unicorns (TC+)

We’re already halfway into 2023, which means we’re only a couple weeks away from brand new, sizzling data for the second quarter. However, it’s always wise to keep an eye on the horizon, so we’ve decided to draw the few conclusions about web3 and unicorn funding trends that we can from early data on the past three months.

Coinbase execs: As global crypto policy grows, U.S. has urgent need for legislation (TC+)

Coinbase, one of the largest crypto exchanges globally, has been around for 10 years. And while the company has grown its offerings, products and services, its policy talking points haven’t changed dramatically, Kara Calvert, head of U.S. policy at Coinbase told TechCrunch+. But what has changed, she said, is the “momentum and urgency” for digital asset legislation and rules at a federal level in the U.S.

AI and crypto integration is going to happen whether you want it or not (TC+)

As artificial intelligence continues to grow to new heights of popularity, industry players are considering new ways the technology could integrate with crypto and blockchains. During Coinbase’s State of Crypto Summit on Thursday, venture capitalists and AI experts shared their thoughts during a panel on what similarities and differences the industries have and how investors, builders and users can capitalize on it.

Crypto startup Pillow, backed by Accel and Quona, to discontinue all services

Singapore-headquartered Pillow plans to discontinue all its services and app in the coming weeks, it warned customers Friday, citing regulatory uncertainty that has claimed countless other crypto startups in recent quarters. It had raised about $21 million altogether and counted Accel India, Quona Capital, Elevation Capital and Jump Crypto among its backers. Pillow revealed its $18 million Series A funding in October last year.

Terraform Labs founder Do Kwon jailed four months in Montenegro

Another chapter was published in the long and bizarre saga of Terraform Labs’ Do Kwon. The disgraced crypto founder will spend four months in a jail in Montenegro for falsifying official documents. The next step for Kwon is still unclear as both the U.S. and South Korea have been seeking to extradite him over charges in both countries relating to the collapse of Terraform Labs.

The latest pod

For this week’s episode, Jacquelyn interviewed Jack Lu, co-founder and CEO of NFT marketplace Magic Eden. This is his second time on Chain Reaction, but the market has evolved a lot since the last time he came on in August 2022, so we’re excited to have him back!

Before co-founding Magic Eden in 2021, Lu worked as a product manager at Google and a consultant for Boston Consulting Group.

Magic Eden originally began as a Solana-based NFT trading platform, but has expanded its support to other blockchain networks like Polygon, Ethereum and Bitcoin. Today, it has grown into one of the largest NFT marketplaces, with over 8,000 collections, about $3 billion in NFT transactions and 22 million unique monthly visitors. In June 2022, Magic Eden raised $130 million in a Series B round that granted it unicorn status.

We discussed why Magic Eden expanded its support to other blockchains, adding BRC-20 token support to its secondary platform and how the company plans on staying competitive in the constantly changing market.

We also talked about:

• NFT market volatility
• Royalty fees
• Web3 gaming expansion
• Advice for NFT community

Subscribe to Chain Reaction on Apple Podcasts, Spotify or your favorite pod platform to keep up with the latest episodes, and please leave us a review if you like what you hear!

Follow the money

1. Bitpanda’s crypto exchange separated from Bitpanda and secured $33 million
2. Gaming platform Mythical Games raised $37 million in an extended Series C1 round
3. Web3 gaming platform Pixion Games raised $5.5 million
4. AI-powered crypto search engine Kaito raised $5.5 million in a Series A round
5. Startale Labs raised $3.5 million in a seed round for web3 infrastructure for public goods

This list was compiled with information from Messari as well as TechCrunch’s own reporting.

To get a roundup of TechCrunch’s biggest and most important crypto stories delivered to your inbox every Thursday at 12 p.m. PT, subscribe here.

Say goodbye to Q2 and the crypto hacks, scams and rug pulls that came with it by Jacquelyn Melinek originally published on TechCrunch

from https://ift.tt/Yzdc5PW
via Technews

Read More

Niantic lays off 230 employees cancels NBA and Marvel games

Pokémon GO maker Niantic laid off 230 employees today, just one year after it laid off around 90 employees.

During last year’s layoffs, Niantic cancelled four projects, including a Transformers game. Some Niantic games will meet the same fate this time around. After four months in the App Store, Niantic is shutting down NBA All-World; the company will also cancel production on a game based on the Marvel franchise.

“In the wake of the revenue surge we saw during Covid, we grew our headcount and related expenses in order to pursue growth more aggressively,” CEO John Hanke wrote in an email to employees, cross-posted to the company blog.

This has been a common refrain among the hundreds of tech companies that have conducted layoffs over the last year; companies claimed they overhired during the pandemic and now need to right-size their teams. In Niantic’s case, Hanke said that revenue has returned to pre-pandemic levels, and new projects have not delivered as much revenue as they would have hoped.

One such new project is Peridot, a Tamagotchi-like mobile game. Niantic’s first attempt at original IP since Ingress, Peridot launched in May. But according to Market Intelligence Firm Sensor Tower, Peridot has only made $1.4 million in gross in-app purchase revenue thus far.

Peridot is a technologically sophisticated game, complete with a robust breeding system that makes every players’ pets genetically unique. But players were disappointed at launch, since many of the game’s most exciting features are paywalled. The only way to hatch a new Peridot, for example, is to pay $5 for an in-game item. And once you pay to hatch a Peridot, you discover that you also have to pay if you want your new pet to have certain unique fur patterns or styles.

Pokémon GO is Niantic’s cash cow, pulling in more than $1 billion in in-app purchases each year since 2020. But players have also been feeling slighted by Niantic’s in-app purchase system.

At the end of March, Niantic nearly doubled the price of remote raid passes, an extremely popular in-app item. Niantic’s reasoning is that remote play options were essential during pandemic lockdowns, but they go against Niantic’s vision for the game, which is to get people outside to play together in person. Players don’t feel that way, though. Some even chose to boycott Pokémon GO in protest. While their protest may not have caught Niantic’s attention, its in-app purchase sales likely did.

Between July 2022 and March 2023, SensorTower data shows Niantic made an average of $70 million in gross in-app purchase revenue each month. In the three months since the remote raid pass price increase, Niantic has made an average of $53 million per month.

“The top priority is to keep Pokémon GO healthy and growing as a forever game,” Hanke wrote. Some Pokémon GO staff were impacted by these layoffs.

Though games like Peridot have not yet proved financially sustainable, Niantic has an entire business arm separate from its own games. Niantic’s Lightship AR developer kit makes it possible for any developer who knows how to use Unity to make AR games. Developers also have access to Niantic’s impressive visual positioning system (VPS), which lets users interact with local landmarks in their real-world surroundings.

Hanke even mentioned in his note to employees that the company wants to ramp up its focus on building for mixed reality devices and AR glasses. So, if Niantic can’t seem to make a successful follow-up to Pokémon GO, maybe it’s developer tools can keep the company on the right track.

The Niantic EC-1

Niantic lays off 230 employees, cancels NBA and Marvel games by Amanda Silberling originally published on TechCrunch

from https://ift.tt/ImMtFdf
via Technews

Read More

Millions affected by MOVEit mass-hacks as list of casualties continues to grow

Hackers have compromised the personal data of more than 15 million individuals by exploiting a security vulnerability in the MOVEit file transfer tool, as the number of victim organizations continues to grow.

There are more than 140 known victims of Clop ransomware attacks targeting a vulnerability in MOVEit Transfer, an enterprise file transfer tool developed by Progress Software. Brett Callow, a ransomware expert and threat analyst at Emsisoft, tells TechCrunch that while only 10 of these victims have so far confirmed the number of people affected, the number already exceeds more than 15.5 million individuals.

This includes approximately 3.5 million Oregon driver license holders; roughly six million Louisiana residents; some 770,00 members of the California Public Employees’ Retirement System; between 2.5 and 2.7 million Genworth Finance clients; approximately 1.5 million customers of insurance provider Wilton Reassurance; more than 170,000 beneficiaries of the Tennessee Consolidated Retirement System; and more than half a million Talcott Resolution customers.

Callow tells TechCrunch that the mass-hacks include U.S. educational non-profit National Student Clearinghouse, which could be a “potentially significant” breach in terms of numbers. The organization, which began notifying schools of the data breach, works with 3,600 colleges and universities and 22,000 high schools.

Callow noted that at least seven of the known MOVEit victims are U.S. universities, and 16 are U.S. public sector organizations.

This includes the U.S. Department of Health and Human Services (HHS), according to Bloomberg, which reported Wednesday that officials notified Congress of an incident involving the exposure of more than 100,000 individuals. HHS did not respond to TechCrunch’s questions and has not yet been added to Clop’s dark web leak site.

U.S. cybersecurity agency CISA previously told TechCrunch that “several” U.S. government agencies had experienced intrusions related to the exploitation of the MOVEit transfer flaw, and a spokesperson for the Department of Energy confirmed that this included two DOE entities.

It’s not just government departments that have been targeted.

Clop, which claimed responsibility for the widespread attacks, has added tens of new victims to its leak site this week alone, including banks, consultancy and legal companies, and energy giants.

Siemens Energy spokesperson Claudia Nehring confirmed to TechCrunch that the company is among the targets of the MOVEit attacks. “Based on the current analysis no critical data has been compromised and our operations have not been affected. We took immediate action when we learned about the incident,” Nehring added.

The University of California (UCLA), which used MOVEit Transfer to transfer files across campus and to other entities, is also among Clop’s newly listed victims. UCLA spokesperson Marge Grey told TechCrunch that the university “notified the FBI and worked with external cybersecurity experts to investigate the matter” and has notified those who have been impacted. UCLA declined to say how many individuals had been affected.

None of the other victims listed by Clop have yet responded to TechCrunch’s requests for comment.

The exact number of impacted organizations, and subsequently breached individuals, remains unknown. In a post on its leak site, Clop claims to have compromised “hundreds” of organizations, which means that more victims are likely to come to light in the coming days and weeks.

In light of this latest wave of mass attacks, U.S. State Department earlier this month offered a $10 million bounty for information on the Clop ransomware group, a Russia-linked gang that was also responsible for previous mass-attacks exploiting flaws in Fortra’s GoAnywhere file transfer tool and Accellion’s file transfer application.

---

Do you work at an organization that’s affected? Do you have more information you can share? You can contact Carly Page securely on Signal at +441536 853968 and by email. You can also share tips and documents with TechCrunch via SecureDrop.

Millions affected by MOVEit mass-hacks as list of casualties continues to grow by Carly Page originally published on TechCrunch

from https://ift.tt/53mhfSv
via Technews

Read More

ChatGPT prompts: How to optimize for sales marketing writing and more

ChatGPT, OpenAI’s AI-powered chatbot, has taken the world by storm.

Capable of writing emails, essays and more given a few short prompts, ChatGPT has become one of the fastest-growing apps in history. Beyond that, it’s begun to find a place in the enterprise, particularly with the launch of plugins that connect the chatbot to third-party apps, websites and services. Most recently, ChatGPT Plus subscribers now have access to a new feature called Browsing, which allows ChatGPT to search Bing for answers to prompts and questions.

But ChatGPT isn’t always the most cooperative assistant. Getting it to output something specific requires careful fine-tuning of the prompts.

A number of resources and guides for ChatGPT prompt writing have sprung up since the tool’s launch. But not all of them are especially easy to follow — or intuitive. To help folks both new to ChatGPT and looking to learn new tricks, we’ve compiled a list of the best ChatGPT prompts for different types of workflows — specifically writing, marketing, sales, students and tech enthusiasts.

The best ChatGPT prompts for sales

No one likes to write sales emails. No one. And while there’s plenty in the way of tools to tackle the task, many rely on templates with inflexible, repetitive language. Not so with ChatGPT.

When writing sales prompts for ChatGPT, though, the wording really matters. For example, consider the prompt:

Write a concise and informal cold email to a sales lead.

Compare it to:

Write a cold email to a sales lead.

You’ll notice that the results for the first, far more descriptive prompt are better — objectively better — than the results for the second. While not perfect, they’re a much better starting point for something, well, sendable.

You can take the ChatGPT prompt fine-tuning further. Let’s say you want copy for LinkedIn prospecting emails — LinkedIn being a great place to look for sales leads (as many marketers know). Try a prompt like:

John’s Linkedin summary: [insert text here] Write a cold email to Katie, who I just found on LinkedIn.

Katie Paterson over at the Zapier blog gave it a shot. The result was impressively personalized — and a lot better than most of the sales spam I’ve gotten over the years, truth be told.

ChatGPT needn’t be confined to the email realm. Vidyard writes about how the tool can be used to automate cold call scripts or sales pitch processes. Try something like:

Write a sales pitch for a marketing consultant offering solutions to small businesses struggling with low online visibility and poor search engine rankings.

Again, you’ll most likely have to tweak the results. But undeniably, it’s a time saver.

The best ChatGPT prompts for marketing

ChatGPT is an excellent marketing tool — or can be, if you use the right set of prompts. As with writing, it requires knowing in which specific ways to prompt the model so that it understands your intention.

As any online marketer knows, keywords are an important part of the puzzle. Fortunately, ChatGPT’s a competent keyword generator. Use the prompt:

Generate a list of keywords for [insert text here], including long-tail and high-performing keywords.

That’ll provide a decent starting prompt for whatever copy you’re trying to write.

Speaking of brainstorming copy, it’s no secret that ChatGPT can come in handy here, too — whether it’s for an ad or social media post. For example, take a look at this prompt from Tory Wenger over at Madgicx, which really illustrates the degree of specificity ChatGPT will accept:

Craft a compelling ad copy for our Facebook ad campaign, targeting users who have previously visited our website and creating a sense of urgency, as well as adding our offer for exclusive promotion to entice them to take action. The offer is [insert text here].

It’s as easy as that.

ChatGPT can also give marketing and brand advice, believe it or not, answering tough questions with surprising depth and nuance. Wordstream’s Gordon Donnelly asked ChatGPT how to respond to negative comments and publicity:

As a social media marketing manager, how do I respond to people that are writing negative things about my products on Twitter?

ChatGPT’s response? A diplomatically-worded email asking for feedback on a product, using wording like “Your feedback is essential to us” and “we want to make sure we’re exceeding your expectations.” Talk about measured!

The best ChatGPT prompts for writing

When it comes to writing, ChatGPT can be a useful companion indeed — serving as a brainstorming tool or streamlining the more monotonous bits of the writing process. But the chatbot isn’t always the most steerable or predictable unless you use very specific prompt wording.

For example, “priming” ChatGPT can set the tone and context. Try a prompt like:

“I’m a tech blogger and I need your help writing a blog post. The topic is CES. This post should be helpful for people who are interested in new and upcoming smartphones. Do not start writing yet. Do you understand?”

That’ll “ground” the tool, providing ChatGPT context for future questions.

Another nifty tip is using bullet points to guide ChatGPT as it writes. Try using a prompt like:

Write an introduction based on the bullet points below:

• This is an article about a new tech product — a wireless air fryer.
• The product costs $20.
• The product will be available for sales on June 16

Given a moment, ChatGPT will generate something coherent that incorporates details from each of the bullets.

ChatGPT can also be “taught” to mimic style, voice and tone — a useful feature in instances where you’re trying to have it complete parts of an article or essay. Trying entering this prompt:

Analyze the text below for style, voice and, tone. Create a prompt to write a new paragraph in the same style, voice and tone. [insert text here]

It might not always get it right. But when instructed to write this way, ChatGPT is much more likely to produce something usable — and insightful.

The best ChatGPT prompts for students

Not every academic institution is on board with the idea of using ChatGPT as a writing tool — or even writing aid. But others are — and have gone to great lengths to incorporate ChatGPT into their curriculums. This writer supports the latter camp, but would advise students against using ChatGPT where prohibited by an instructor. You’ve been warned.

The sky’s the limit, really, when it comes to education-focused ChatGPT prompts. It really depends on the task at hand and the nature of the work. You could try, for instance, a prompt like this:

Help me write a research paper on the causes of the American Revolution

Or a prompt like:

Can you help me explain the significance of the Magna Carta?

And ChatGPT will do its best to respond in a way that makes sense — if not perfect sense.

A word of warning when asking ChatGPT for facts and figures: It doesn’t always get it right. Sometimes, thanks to a phenomenon known as hallucination, the chatbot invents things — very confidently — out of whole cloth. That’s why it’s wise to fact-check answers from ChatGPT before pasting them into a piece.

Once again, ChatGPT can be asked to do more than simply write an essay or answer basic topical questions. Consider this prompt:

Help me create a study plan for my upcoming exams in history and political science.

You’ll need to be more specific than “history and political science,” lest the advice be overly broad. But ChatGPT — while it won’t do the studying for you — should provide a reasonable starting point.

The best ChatGPT prompts for tech enthusiasts

We’ve established that ChatGPT is a fine writer. But did you know that it’s a coder, too, and a mathematician?

Say you want to create a basic web form to collect contact information. ChatGPT will happily do that for you with a prompt like:

Act as a JavaScript Developer, Write a program that checks the information on a form. Name and email are required, but address and age are not.

The resulting code may contain some mistakes. ChatGPT certainly isn’t perfect. But it should be a reasonable starting point.

In a more sophisticated use case, ChatGPT can write database queries for applications — a task that normally takes a fair amount of time (and, sometimes, trial and error). Try this prompt for MySQL, one of the more popular relational database systems:

Write a MySQL Query

Tables: users and orders

Requirement: It should give user details who placed highest order today

Again, the results won’t be usable out of the box necessarily. But they’ll help you to get where you need to be.

Same goes for math questions. One of my favorite recent prompts from PromptHero, an AI prompt database, is this:

I want you to act as a math teacher. I will provide some mathematical equations or concepts, and it will be your job to explain them in easy-to-understand terms. This could include providing step-by-step instructions for solving a problem, demonstrating various techniques with visuals or suggesting online resources for further study. My first request is “I need help understanding how probability works.”

ChatGPT, you’ll find, can be a surprisingly thoughtful tutor.

ChatGPT prompts: How to optimize for sales, marketing, writing, and more by Kyle Wiggers originally published on TechCrunch

from https://ift.tt/DNsAV0M
via Technews

Read More

As $100M venture rounds evaporate IPOs might have to carry the weight

Earlier this year we wrote that the “the $100 million venture round is going extinct.” Often our predictions wind up sideways. This time we were on the right track.

According to new data from PitchBook, the U.S. venture market is continuing to endure lackluster velocity for nine-figure investments into private companies, colloquially referred to as “mega-rounds.”

---

The Exchange explores startups, markets and money.

Read it every morning on TechCrunch+ or get The Exchange newsletter every Saturday.

---

In the first half of 2023, PitchBook counted just 108 mega-rounds in the United States. If we presumed that this rate will hold throughout the year, we’re looking at just over 200 nine-figure deals in the U.S. in 2023. That’s a dramatic decline from prior levels. Starting in Q4 2020 through Q3 2022, there were more than 100 mega-rounds recorded per quarter. In 2021, the average was more than 200 per quarter. To see perhaps 200 this year implies that the number of late-stage startups that will be able to raise an IPO-sized round is in free fall.

The rounds are also getting smaller, with data indicating that the average nine-figure round size has fallen under the $200 million mark, exclusive of a few rounds that are hardly traditional venture deals, like OpenAI’s massive round earlier this year. Smaller mega-rounds, and fewer of them, is a tough mix for unicorns of all stripes and sizes.

Of course, we could see nine-figure rounds rebound in other markets. Europe and Asia have seen their fair share of the transactions historically. But as the United States’ venture market is the largest in the world and was once the leading player in mega-round financings, where the U.S. goes, so, too, goes the world.

Cava’s explosive IPO hints that public markets are more ready for growth-stories than founders thought

If unicorns here are struggling to find fodder in the quantity that they became accustomed to, other startups around the world are likely enduring a similar dearth of capital.

Notably PitchBook thinks that “the need for capital likely leading to an uptick in mega-rounds as the year progresses” thanks to “the notion that depleting cash runways will force more of these startups to raise in the harsher dealmaking environment,” it still expects full-year mega-round tallies to come in at dramatically reduced levels compared to prior years.

As $100M+ venture rounds evaporate, IPOs might have to carry the weight by Alex Wilhelm originally published on TechCrunch

from https://ift.tt/V6HrxgM
via Technews

Read More

Techstars raising $150 million for new accelerator fund

Startup accelerator Techstars is raising $150 million for its new fund, SEC filings show. The accelerator is seeking new capital ahead of the end of the deployment period of its third institutionally-backed fund, Techstars Accelerator 2021, later this year.

Techstars declined to offer details on how it intends to use or deploy the funding, citing regulatory restrictions. However, like its predecessors, the new fund, Techstars Accelerator 2024, is expected to be used for accelerator-stage and/or post-accelerator investments.

Notably, the Boulder-based early-stage investor has used its current fund only for accelerator-stage investments. Its previous funds backed startups both during the accelerator stage and with follow-on investments after the program.

Fund 3 aimed to back over 800 pre-seed and seed startups that are part of its global accelerator programs by the end of 2023. Through its accelerator groups, Techstars has so far backed over 3,500 early-stage startups in various sectors including healthtech, fintech, web3 and cleantech.

Startups going through Techstars’ three-month program receive $20,000 and a $100,000 convertible note in exchange for 6% to 9% of common stock, in addition to access to its network and mentorship, amongst other resources.

Techstars debuts new fund for companies too early for its own accelerator

Founded in 2006, Techstars runs over 50 accelerator groups in major cities across the globe, including New York, Los Angeles, London, Boston, Tel Aviv and Lagos. The accelerator launched the Lagos arm last year in partnership with ARM Labs as part of its plan to expand its specialist and generalist programs globally.

To double down on its growth and expansion plans, Techstars recently launched new funds to bolster its investment efforts.

Last year, for instance, it launched Rising Stars, a fund to back startups that are too early for its accelerator programs. Under this fund, Techstars will issue $100,000 checks in exchange for 7% to 10% equity.

The accelerator also launched the $80 million Advancing Cities Fund in partnership with JPMorgan to back over 400 startups by diverse founders in nine cities across the U.S.

Techstars names Maëlle Gavet CEO as the accelerator group looks to expand

Techstars raising $150 million for new accelerator fund by Annie Njanja originally published on TechCrunch

from https://ift.tt/oic1kaZ
via Technews

Read More

OpenAI brings the competition to DeepMinds doorstep with new London office

OpenAI is expanding overseas. To London, specifically.

Today, the Microsoft-backed AI startup announced that it plans to open an office in London, its first international outpost. When OpenAI’s London location opens its doors, it’ll focus on advancing “research and engineering capabilities” while balancing collaborating with “local communities and policymakers,” according to CEO Sam Altman.

“We see this expansion as an opportunity to attract world-class talent and drive innovation in AGI development and policy,” Altman, who reportedly had floated Poland and France as alternatives for the office, said in a canned statement. “We’re excited about what the future holds and to see the contributions our London office will make towards building and deploying safe AI.”

London is a conspicuous choice for OpenAI, which hasn’t expanded beyond its San Francisco headquarters since its founding in 2015. The city is the longtime home base of DeepMind, Google’s largest AI research division, and a wellspring of data science talent, owing to its rich academic history and renowned universities.

Broadly speaking, London is also becoming a booming center for AI startup ventures. According to a recent report, as of 2021, over 1,300 AI companies were based in London and the city was the top-funded in the U.K. in terms of venture dollars invested.

The city is also important politically to tech companies heavily invested in AI, like OpenAI, who seek to convince the U.K.’s governing bodies to regulate AI with a light touch. On a recent lobbying tour, Altman made an appearance at the University College London, where he called for “balanced” regulation and warned of the risks of deepfake disinformation.

At that same appearance, Altman said that OpenAI would “cease operating” in the European Union if it’s unable to comply with the provisions of the bloc’s AI Act, one of the first comprehensive set of regulations for the AI industry. He later backed down from the comments — but the play was made.

OpenAI brings the competition to DeepMind’s doorstep with new London office by Kyle Wiggers originally published on TechCrunch

from https://ift.tt/GgUiZaJ
via Technews

Read More

Vertical AI: The next logical iteration of vertical SaaS

Paris Heymann Contributor

Paris Heymann is a partner at Index Ventures, where he invests primarily in B2B SaaS and data.

More posts by this contributor

• Acquisition, retention, expansion: Why SaaS founders must understand GDR and NDR
• Metrics that matter: 3 KPIs to track on the path to profitability

At Index Ventures, we view the emergence of vertical SaaS (vSaaS) — cloud-based software tailor-made for specific industries — as part of a broader trend of end users increasingly demanding superior technology products.

Consumers want solutions-oriented software made specifically to solve their exact business problems. In an environment where we are inundated with software, narrow and specific is well-positioned versus broad and generalized.

The concept is not new: Even the largest horizontal tech companies verticalize their sales organizations and product features when they have enough scale within each vertical for that to be a sensible approach.

Cloud giants AWS, Azure, and Google Cloud Platform prominently feature vertical industry solutions with dedicated sales teams, as do other large platforms like Salesforce, ServiceNow, Snowflake and Workday.

These tech leaders verticalize their offerings over time because it’s a high-quality experience for customers and end users when a technology vendor deeply understands the industry, has sales and support reps attending the same conferences as users, and is rapidly evolving the product to suit customer needs.

The AI category is rapidly evolving, but developing into three layers: foundational models, AI infrastructure, and AI applications.

With the AI platform shift upon us, we believe that the next logical iteration of vertical SaaS will be vertical AI – vertically-focused AI platforms, bundled alongside workflow SaaS, built on top of models which have been uniquely trained on industry-specific datasets.

Why vertical AI?

The AI category is rapidly evolving, but developing into three layers: foundational models, AI infrastructure, and AI applications.

Examples of AI stack startups. (Index Ventures is an investor in Causaly, Cohere, Scale, ServiceTitan and Weaviate.) Image Credits: Index Ventures

Foundational models are the bedrock of the AI stack. Leaders in this space include Anthropic, Cohere, and OpenAI. It’s likely there will be a limited number of vendors in the foundational LLM space given the high capital requirements to build and train models.

The “picks and shovels” of AI sit at the infrastructure layer, a catch-all which includes a variety of categories including data enhancement, fine-tuning, databases, and model training tools. For example, vector databases like Pinecone and Weaviate are gaining significant adoption.

Other companies like Scale are being used for data generation, labeling, and training. Hugging Face has emerged as a leader for model discovery and inference. Weights & Biases is widely recognized within MLOps. LangChain is an open-source development framework used to simplify the creation of new applications using LLMs. These are a few of many companies which are helping companies transform models and data into products.

Foundational models and infrastructure are enabling an explosion of AI business applications. These AI-powered applications could be used by any end user, in any industry, to accomplish an array of tasks.

Vertical AI: The next logical iteration of vertical SaaS by Walter Thompson originally published on TechCrunch

from https://ift.tt/mh7vNSP
via Technews

Read More

Reka emerges from stealth to build custom AI models for the enterprise

Large language models (LLMs) like OpenAI’s GPT-4 are all the rage these days, owing to their unparalleled ability to analyze and generate text. But for organizations looking to leverage LLMs for specific tasks — say, writing ad copy in a brand’s style — their generalist nature can become a liability.

When the instructions get too precise, even the best LLMs struggle with consistency. Fine-tuning, or narrowing an LLM’s scope, is one solution. But it’s often challenging from a technical standpoint, not to mention costly.

Motivated to find an easier way, a team of researchers from DeepMind, Google, Baidu and Meta founded Reka, which emerged from stealth today with $58 million. DST Global Partners and Radical Ventures led the tranche with participation from strategic partner Snowflake Ventures, alongside a cohort of angel investors that included former GitHub CEO Nat Friedman.

San Francisco-based Reka is the brainchild of Dani Yogatama, Cyprien de Masson, Qi Liu Head and Yi Tay. While working on AI systems including DeepMind’s AlphaCode and Bard, they four co-founders say that they realized it was impractical to expect a large LLM to be deployed for all possible use cases.

“We understand the transformative power of AI and would like to bring the benefits of this technology to the world in a responsible way,” Yogatama told TechCrunch in an email interview. “Reka is a research and product company that develops models to benefit humanity, organizations and enterprises.”

Reka’s first commercial product, Yasa, doesn’t quite meet those lofty ambitions. But it exemplifies the startup’s early approach. Going beyond text, Yasa is a multimodal AI “assistant” trained to understand images, videos and tabular data in addition to words and phrases. It can be used to generate ideas and answer basic questions, Yogatama says, as well as derive insights from a company’s internal data.

In this way, Yasa, which is in closed beta, isn’t dissimilar to models like GPT-4, which can also understand text and images. But the twist is that Yasa can be easily personalized to proprietary data and applications.

“Our technology allows enterprises to benefit from progress in LLMs in a way that satisfies their deployment constraints without requiring a team of in-house expert AI engineers,” Yogatama said.

Yasa is just the start. Next, Reka plans to turn its attention to AI that can accept and generate even more types of data and continuously self-improve, staying up to date without the need for retraining.

To that end, only available to select customers for now, Reka also provides a service to adapt LLMs it developed to custom or proprietary company data sets. Customers can run the “distilled” models on their own infrastructure or via Reka’s API, depending on the application and project constraints.

Reka, it should be noted, isn’t the only startup chasing after models better suited for enterprise use cases. Writer lets customers fine-tune LLMs on their own content and style guides. Contextual AI and LlamaIndex, which recently emerged from stealth, are developing tools to allow companies to add their own data to existing LLMs. And Cohere trains LLMs to customers’ specifications.

Not to be outdone, incumbents like OpenAI now offer tools for fine-tuning models and connecting them to the internet and other sources to ensure that they remain up to date.

But Reka’s sales pitch won over one early customer (and investor), Snowflake, which partnered with the startup to let Snowflake customers deploy Yasa from their accounts. Appen, the big data analytics company, also recently announced that it’s working with Reka to build tailored multimodal model-powered apps for the enterprise.

Rob Toews, a partner at Radical Ventures, had this to say when asked why he invested in Reka:

“What makes Reka unique is how it offers every business the power and potential of an LLM without having to put up with many tradeoffs,” Toews said via email. “Reka’s distilled Yasa models keep the data within the enterprise, they’re incredibly efficient in terms of cost and energy and they don’t require costly research teams building models from scratch. If every business will become an ‘AI’ business, Reka’s ambition is to give each of those businesses its own, production-quality foundation model.”

Yogatama says Reka, which currently isn’t generating revenue, will use its funding to date to acquire computing power from Nvidia and build a business team.

Reka emerges from stealth to build custom AI models for the enterprise by Kyle Wiggers originally published on TechCrunch

from https://ift.tt/nvW0Jwu
via Technews

Read More

How Sheins influencer trip to a Chinese factory backfired

As Shein eyes an IPO, the company’s image needs a serious makeover. From stealing indie designers’ work, to violating local labor laws, Shein has fallen out of vogue on social media, so the company invited a group of influencers to tour one of its factories in Guangzhou, China.

“I expected the facility to be so filled with people just slaving away, but I was actually pleasantly surprised that most of these things were robotic,” said Destene (@itsdestene_), a creator with more than 4 million followers, in a TikTok. “Honestly, everyone was just working like normal, like chill, sitting down, they weren’t even sweating.”

In a now-deleted video, fashion influencer Dani (@itsdanidmc) reflected on her trip with Shein by saying that “there is a narrative fed to us in the U.S” about Shein, but that her biggest takeaway from the trip was to be an “independent thinker, get the facts, and see it with your own two eyes.”

As commenters were quick to point out on these influencers’ posts, it’s difficult to believe that what we see in these videos reflects the reality of Shein factory working conditions. Rather, this was a highly curated brand trip wherein influencers are offered free travel opportunities and gifts, encouraging them to promote a favorable image of the company.

“It feels like they used you for damage control and it’s disturbing,” one commenter wrote on Destene’s video.

“If they wanted to really show they ain’t on nothing they’d invite investigative journalists, not influencers they can pay off lol,” another commenter said.

A TikToker who posts about sustainable fashion, Ella (@myweeklyyarn) spoofed the ordeal by making a video pretending to tour the Triangle Shirtwaist Factory, which was the site of one of the deadliest industrial fires in U.S. history.

“Surely this collab will have no consequences for me or any of the real #sheinfactory workers,” Ella wrote in the TikTok’s caption.

The Chinese fast fashion giant Shein has become one of fastest growing e-commerce companies in the world; according to TIME, the company did $100 billion in sales in 2022, up from $10 billion in 2020. With low prices, trendy items and a wide range of sizes, Shein has surpassed Amazon on the App Store, ranking #2 on the Shopping charts and #14 overall for free apps.

Amid Shein’s ascent to international dominance, customers have been skeptical — if a company can add thousands of new items each day while keeping prices shockingly low, how can it be operating ethically? While Shein haul videos are abundant on TikTok, Instagram and YouTube, more creators have spoken out against the company for exemplifying the worst of fast fashion’s environmental impact. A CBC investigation found that some Shein items contained significant amounts of lead — one toddler jacket had almost 20 times the amount of lead that Canadian health officials consider safe for kids.

“Shein is committed to transparency and this trip reflects one way in which we are listening to feedback, providing an opportunity to show a group of influencers how Shein works through a visit to our innovation center and enabling them to share their own insights with their followers,” Shein said in an emailed statement. “Their social media videos and commentary are authentic, and we respect and stand by each influencer’s perspective and voice on their experience.”

As the backlash against these influencers hit a crescendo, Dani DMC went on Instagram to respond to critics. She said that when she went on a trip to Lake Tahoe with Shein a few months ago, she relayed questions from her fans about Shein’s labor practices to higher-ups at the company, who gave her “what felt like authentic” answers.

“[Shein] brought this China trip to my attention, and they’re like, ‘You know, we’re aware of all these rumors and all this stuff that’s going on, and we want to put an end to it,'” Dani said in the video. “We want to put our money where our mouth is, show you what’s going on, and for me to be confident in having a future with them or working with them.”

As a plus size model, Dani said she was excited to be working with Shein, since they are one of the only brands selling affordable and trendy plus size clothing. But she admitted that she didn’t do enough background research before going to China with Shein.

“This whole experience has caused me to reevaluate myself, my brand, and to fight even harder for sustainable options for plus size people, and to just be so much more particular with who I’m working with,” she said.

Shein valuation reportedly plummets by a third as it seeks $3B

Why Amazon should pay attention to Shein

How Shein’s influencer trip to a Chinese factory backfired by Amanda Silberling originally published on TechCrunch

from https://ift.tt/Fuy2GiB
via Technews

Read More