by tyler | Oct 21, 2022 | Pulse, Tech
Clearview AI, the controversial facial recognition firm that scrapes selfies and other personal data off the Internet without consent to feed an AI-powered identity-matching service it sells to law enforcement and others, has been hit with another fine in Europe.
This one comes after it failed to respond to an order last year from the CNIL, France’s privacy watchdog, to stop its unlawful processing of French citizens’ information and delete their data.
Clearview responded to that order by, well, ghosting the regulator — thereby adding a third GDPR breach (non-cooperation with the regulator) to its earlier tally .
Here’s the CNIL’s summary of Clearview’s breaches:
“Clearview AI had two months to comply with the injunctions formulated in the formal notice and to justify them to the CNIL. However, it did not provide any response to this formal notice ,” the CNIL wrote in a press release today announcing the sanction [emphasis its].
“The chair of the CNIL therefore decided to refer the matter to the restricted committee, which is in charge for issuing sanctions. On the basis of the information brought to its attention, the restricted committee decided to impose a maximum financial penalty of 20 million euros , according to article 83 of the GDPR [General Data Protection Regulation].”
The EU’s GDPR allows for penalties of up to 4% of a firm’s worldwide annual revenue for the most serious infringements — or €20 million, whichever is higher. But the CNIL’s press release makes clear it’s imposing the maximum amount it possibly can here.
Whether France will see a penny of this money from Clearview remains an open question, however.
The U.S.-based privacy-stripper has been issued with a slew of penalties by other data protection agencies across Europe in recent months, including €20M fines from Italy and Greece ; and a smaller U.K. penalty . But it’s not clear it’s handed over any money to any of these authorities — and they have limited resources (and legal means) to try to pursue Clearview for payment outside their own borders.
So the GDPR penalties look mostly like a warning to stay away from Europe.
Clearview’s PR agency, LakPR Group, sent us this statement following the CNIL’s sanction — which it attributed to CEO Hoan Ton-That:
There is no way to determine if a person has French citizenship, purely from a public photo from the internet, and therefore it is impossible to delete data from French residents. Clearview AI only collects publicly available information from the internet, just like any other search engine like Google, Bing or DuckDuckGo.
The statement goes on to reiterate earlier claims by Clearview that it does not have a place of business in France or in the EU, nor undertake any activities that would “otherwise mean it is subject to the GDPR”, as it puts it — adding: “ Clearview AI’s database of publicly available images is lawfully collected, just like any other search engine like Google.”
(NB: On paper the GDPR has extraterritorial reach so its former arguments are meaningless, while its claim it’s not doing anything that would make it subject to the GDPR looks absurd given its amassed a database of over 20 billion images worldwide and Europe is, er, part of Planet Earth… )
Ton-That’s statement also repeats a much-trotted out claim in Clearview’s public statements responding to the flow of regulatory sanctions its business attracts that it created its facial recognition tech with “the purpose of helping to make communities safer and assisting law enforcement in solving heinous crimes against children, seniors and other victims of unscrupulous acts” — not to cash in by unlawfully exploiting people’s privacy — not that, in any case, having a ‘pure’ motive would make any difference to its requirement, under European law, to have a valid legal basis to process people’s data in the first place.
“We only collect public data from the open internet and comply with all standards of privacy and law. I am heartbroken by the misinterpretation by some in France, where we do no business, of Clearview AI’s technology to society. My intentions and those of my company have always been to help communities and their people to live better, safer lives,” concludes Clearview’s PR.
Each time it has received a sanction from an international regulator it’s done the same thing: Denying it has committed any breach and refuted the foreign body has any jurisdiction over its business — so its strategy for dealing with its own data processing lawlessness appears to be simple non-cooperation with regulators outside the US.
Obviously this only works if you plan for your execs/senior personnel to never set foot in the territories where your business is under sanction and abandon any notion of selling the sanctioned service to overseas customers. ( Last year Sweden’s data protection watchdog also fined a local police authority for unlawful use of Clearview — so European regulators can act to clamp down on any local demand too, if required.)
On home turf, Clearview has finally had to face up to some legal red lines recently.
Earlier this year it agreed to settle a lawsuit that had accused it of running afoul of an Illinois law banning the use of individuals’ biometric data without consent. The settlement included Clearview agreeing to some limits on its ability to sell its software to most U.S. companies but it still trumpeted the outcome as a “huge win” — claiming it would be able to circumvent the ruling by selling its algorithm (rather than access to its database) — to private companies in the U.S.
The need to empower regulators so they can order the deletion (or market withdrawal) of algorithms trained on unlawfully processed data does look like an important upgrade to their toolboxes if we’re to avoid an AI-fuelled dystopia.
And it just so happens that the EU’s incoming AI Act may contain such a power , per legal analysis of the proposed framework.
The bloc has also more recently presented a plan for an AI Liability Directive which it wants to encourage compliance with the broader AI Act — by linking compliance to a reduced risk that AI model makers, deployers, users etc can be successfully sued if their products case a range of harms, including to people’s privacy.
Clearview AI banned from selling its facial recognition software to most US companies
France latest to slap Clearview AI with order to delete data
by tyler | Oct 19, 2022 | Pulse, Tech
“Both my dad and mom had their own business,” Ally Robotics founder and CEO Mitch Tolson tells TechCrunch. “My mom had a sign company. Every single weekend and nights during the week, I was installing neon signs, welding up frames, digging trenches, holes for electrical, all of it.”
The executive says those formative years on construction sites were part of the inspiration behind the Washington-based startup’s formation. The central conceit behind the company is simple enough: What if we could train robots more like we train people?
The company, which pitched today onstage in Startup Battlefield at TechCrunch Disrupt, has developed a combination hardware and software solution designed to make it easier to deploy these automated solutions for those without coding/robotics experience. The perfect example is Ally’s first major partner, Miso Robotics. The company behind the Flippy robotic fry chef signed a $30 million letter of intent to adopt Ally’s technology. That’s a big win at this early stage.
The startup has also raised $4.7 million in crowdfunding, along with a $6.1 million Series A led by Joe Rodriguez of Pancho Ryan LLC. The company is currently in development for its robot arms, with plans to begin manufacturing late next year. Beyond the already announced Miso deal, the company has additional letters of intent, amounting to around $200,000/month, in a kind of RaaS (robotics-as-a-service) model.
Beyond flipping burgers, Tolson says construction workers — including a roofer — have expressed interest in the platform, along with Bobacino, an automated boba tea bar. Yet another firm has expressed interest in the technology to help paint Christmas ornaments.
The company says it’s developed both hardware and software components in tandem to lower the barrier of entry for non-robotics, creating a no-code solution in the process. Users show the robot how to perform a task by walking them through the steps, be it making a hamburger or affixing shingles to a roof.
“I think that to build a great product that really focuses on solving the real customer need and requirements, you can’t just be a software company or a hardware company,” says Tolson. “You have to do both.”
by tyler | Oct 19, 2022 | Pulse, Tech
In a move that broke the hearts of many a tennis fan around the world, Serena Williams recently announced that she would stop playing professional tennis. But today onstage in front of a packed auditorium at TechCrunch Disrupt, she reminded us that while she brings that competitive mindset to running Serena Ventures, an early-stage VC firm, she says that her professional mindset is “quite the opposite” of her investor persona.
“I think that’s a different Serena. I’m very serene,” she says. “I’m the one that brings that very calm disposition.”
Just don’t mistake that tranquility for passivity. Williams still brings her killer instinct to investing out of her fund, along with her general founding partner, Allison Rapaport Stillman.
She also makes diversity a non-negotiable. About 40% of her LPs are women and people of color, something that is practically unheard of in the venture industry. Williams also invests – in the pre-seed, seed and Series A stages – with the mindset that financial inclusion is far more than a buzzword.
“ We spend a lot of our time in consumer-facing technology because we want to build products for the communities we want to serve…” she said. “We spend a lot of time on gender and racial equity within the healthcare system. mental health, mental health and wellness. These are problems that are facing every community, but especially facing women and people of color, there’s not enough tools and services there for them. So that’s something we’re really focused on.”
Beyond that, Williams prides herself on authenticity and passion and unsurprisingly, those are characteristics that she looks for in founders, too. According to Stillman, one of Williams’ biggest strengths is being able to read people with just one meeting – being able to tell if a founder really quickly if a founder is someone she wants to get to know better, or not.
One of the questions she asks potential portfolio companies is are they operating because “it’s a white space” or are they doing it “because it’s something near and dear to you.”
“I’ve found that usually when it’s near and dear to you, and there’s a space available for it – that’s the winning combo,” Williams said. “But if you’re doing it just because you see it’s a great opportunity, then no matter how big and open it is, it is generally never as good as someone that really has something that they’re actually passionate about.”
“We like to keep things very authentic,” she insisted. “SV has its own brand. We’re not going to do something that’s not authentic to us or something we don’t believe in.”
She also is curious about more than a founder’s previous successes. She wants to know what founders have not been successful at.
“That shows me that when times get tough, you know what’s coming and you know what to expect and failure is great,” Williams added. “I think it’s a great opportunity to learn and I don’t even like the word failure. I feel like it should be called: ‘I had an opportunity to learn that’s what it should be called to me because I feel like that really creates the very next best thing.”
For the unacquainted, Williams’ husband is venture capitalist Alexis Ohanian – founder of 776 and former co-founder of Reddit. Despite having done a couple of deals together in the past, Williams said the couple doesn’t talk about business at all at home and work to “stay in their own lanes.”
“We are both so competitive,” she said. “I’m internally so competitive. You can’t read my face if inside I’m screaming and dying. We don’t want that so we keep it very separate.”
Speaking of past investments, Williams and Stillman do have a regret or two.
Early on in her investing career, Williams says she passed on a French startup because she “just didn’t have the infrastructure or the time.”
While she wouldn’t name the startup, she did say it was “before people were doing ride-shares.”
But just the fact she saw its potential is validation for Williams: that her instincts were right. Stillman also says the firm passed on investing in the Series B round for delivery giant Instacart – which is now on the verge of going public – because it was a bit late stage for them.
“There was still a lot of room for growth, though,” she laughs.
by tyler | Oct 19, 2022 | Pulse, Tech
The Biden-Harris administration, through the Department of Energy (DOE), announced Wednesday that 20 battery companies will receive a combined $2.8 billion to build and expand commercial-scale facilities in 12 states.
This is the first phase of $7 billion in total from President Biden’s Infrastructure Law that aims to strengthen domestic battery supply chains and reduce reliance on China for battery supply and production as more vehicles become electrified. The companies getting funds will work to extract and process lithium, graphite and other battery materials, manufacture components and demonstrate new ways of acquiring critical materials, including battery recycling, domestically or within free trade agreement countries, the DOE said.
The DOE didn’t specify which companies were awarded funds, but Gene Berdichevksy , co-founder and CEO of battery chemistry company Sila, told TechCrunch his startup received $100 million from the fund. Sila is replacing the graphite in the anode with silicon, which the company says not only makes a better battery but also avoids the supply chain constraints from sourcing graphite from China. The money will go toward Sila’s upcoming Moses Lake, Washington facility , which aims to deliver 10 gigawatt-hours of capacity annually.
Berdichevsky will likely touch on the news today at TechCrunch Disrupt , where he’s joining a panel on building companies with longer time horizons.
The DOE said the federal investment will be matched by recipients to bring in more than $9 billion to boost American production of clean energy technology, which supports Biden’s goal of all new vehicle sales being electric by 2030.
Earlier this year, Biden also passed into law the Inflation Reduction Act, which included a set of stipulations for consumers to qualify for EV tax credits , including the need for critical battery materials to be sourced domestically or in free trade agreement countries, rather than China.
The DOE has a series of goals for the selected projects, including developing enough battery-grade lithium to support about 2 million EVs annually, enough graphite to support 1.2 million EVs annually and enough nickel to supply about 400,000 EVs annually.
The funding will also help install the first large-scale, commercial lithium electrolyte salt production facility in the U.S. and the first lithium iron phosphate cathode facility in the U.S., among other facilities.
The DOE did not respond in time to TechCrunch’s request for comment or more information.
Can battery recycling help end US reliance on China?
by tyler | Oct 19, 2022 | Pulse, Tech
Artists are caught in the crosshair. A world of “we will never get automated” just got automated. In the past few years artwork is now AI Assisted; with simple, open source, and wildly spread code. The world of AI Assisted artwork is a blossoming flower to most people, however there’s a world of theft and ethical practices being consistently misunderstood. Thought leaders are created over night without any identification possible. Will the government ever catchup?
My theory; By creating a system to take advantage of open source models/algorithms that can steal from other artists without their knowledge, I can increase the visibility into unethical practices, and demonstrate that anyone can stylize an image. I grabbed a tutorial from tensorflow, their easiest tutorial, and turned it into a “LinkedIn profile” image input, plus someone’s street artwork, equals my net new image.
My goal; Improve AI assisted artwork ethical practices by demonstrating how unethical it can be. Give end users the ability to see where the problems live by expressing it in text below.
Hello and welcome to this Collab python book (links take you to google free python application stored in their cloud, where I’ve tested this application many times, and it contains my opinions in this top half and below is a complete set of code that will allow you to quickly stylize images of your choosing. If you want to build this environment on your computer, check out my blog on setting up anaconda distribution – a data science toolkit.
What is this tldr content by tyler garrett? I’m talking about fast style transfer for arbitrary styles with beginner notes to help anyone use the solution here with “url” links to two images. You can use your LinkedIn photo, or any image online (most image URLs will work). I leave an image of myself (LinkedIn) and an image of a planet for example material.
We are seeing a more attention focused in AI assisted artwork and I feel obliged to offer some advice/training. I think it’s necessary to say that I feel this is unethical if used incorrectly and at the same time it’s incredible cool because I’ll never in my lifetime create something this mesmerizing in this amount of time.
It is becoming easier than ever to take photos of some ones painting in real life or street artwork, and use it as a style for my profile image on LinkedIn. Below this essay are 4 different photos/examples.
Unethical practices; This is where the gray area presents itself, what if I’m taking images from an ecommerce website and using it to create my own artwork. Their artwork as an input. What if I profit from this artwork?
Where do we draw the line and will anyone notice the differences? When automating style transfers into our content, are we considering ethics or are we throwing it out the door?
I believe we are always building on someone else’s artwork, and there’s nothing truly original or unique. However if I can grab some ones ecommerce store product and turn it into my own art, do you feel I’ve cross a gray area?
The future will be full of scam artists; Creating artwork automatically while pretending the AI didn’t assist, now you’re thought leader-ing you’re a great artist… well this might be the next level of fame, or are we already there today? (We are already seeing it today on social media)
A lot of accounts online today are not images of humans, fake accounts, or what they call “sock puppet accounts.” Who’s to say some of the famous ones aren’t using AI to generate their fame, revenue, and the more you think about it.. Doesn’t sound half bad for a computer to do your work.
If NFT artists using AI artwork as their entire recipe for success, do we care, or are we trying to earn money? Will this small chunk of code be successful at creating money? I don’t know atm, but currently the NFT industry is down 50%.
Questions to ask before buying NFT or using these AI apps.
Numbered below however the moral of the story is “who am I stealing from” and “should I care?”
- Should the original artist be compensated or at a bare minimum notified?
- Will technology catch up to the point of seeing who’s using someone else’s work to create their own or maybe a reverse engineering tool?
- Is that even possible?
- Also, should we start asking ourselves if the artwork we are seeing today is either created by an artist or a machine…
- Does it matter?
It would seem important that we manage this because it’s going to get out of hand but how exactly will we do that? I’ve read plenty of articles about the negative side effects, however the writers aren’t using the technology most of the time and that means they have no feedback on the capability or even flexibility. When you put it that way, maybe we should start consider starting to govern these AI applications prior to generation? Maybe a best practice for AI for ethics reasonings?
Another oddity is people are now asking AI to draw for them. A strange idea to think we would ask a computer to do a hand drawing but that’s what’s happening and that’s starting to scare a lot of artists because they are starting to look more and more realistic. Also, you can say “hand drawn with pencil” and the AI will make it seem like someone actually did this manually..
The dark side; Do you want someone training their art model using images of your children and what would be the damage to you if they did? This is enough to keep me from posting any photos of my children on the internet.
Hello, my name is Tyler Garrett and I’ve been in the enterprise technology solutions realm and business intelligence for the past 12 years. I’m excited to make this ‘AI assisted’ stylization artwork blog because if we make this more accessible for non-technical end users, we will gain more visibility into the ethics concerns that will happen in Ai assisted art world.
Dark Side 2.0; NFTs could be artists stealing someone else’s work and profiting anonymously. There’s no real governance.
The dark side of NFT is anyone can profit infinitely on anyone else’s work. I was interested in this realm because I had an idea that was relevant to stylizing videos for music, however when I got started demoing this application in google collaboratory, I found I was stealing peoples street art and they have no idea.
Then we have open sourced AI assisted artwork being created from text, which is starting to bridge deeper ethics boundaries related to “permission and consent.” People are not giving permission to others, like celebrities are not approving nude images to be created, which is leading down a very devious concern for many parents, celebrities, and people who have the time to plug in and understand what’s happening.
Because of the boom, it feels like an appropriate time to be diligent with the documentation and type more information here than engineers would normally write. The more visibility we can bring to the topic from non-technical people, the better the adoption and chance of improved future algorithms.
Computer vision technology will expand because smart people like to share the things they are developing and even if it can be an ethics concerns they are not thinking about this rather they are often thinking of growing their following by being a creator of something they believe is technical and cool. Not all creators are like this but most enjoy the attention. I don’t mean to be rude, I too fall victim to the gamification of social media platforms and would love to be recognized for creating cool tech.
Engineering at these high levels means you normally don’t explain much and expect everyone to understand or read documentation written in the same difficult method.
I feel it’s important someone simplifies the logic for people who are not engineers because we need the feedback from people who are not engineers too. The faster everyone catches up to the small percent of people who understand computer vision, the faster the laws will catch up, and this will increase adoption of the usage of AI assisted artwork in a positive way (i hope) OR the removal of bad actors, quickly.
Engineers who understand this technology are very rare and likely .001% of the population or less. That means a very small group of people will be able to suggest the ethics are right or wrong. We need more visibility into this topics and thus I’ve created this Fast Style Transfer for Arbitrary Styles for Beginners & notes from Tyler Garrett.
by tyler | Oct 11, 2022 | Pulse, Tech
At Meta Connect 2022 , the company’s annual developer conference for its VR efforts and Oculus hardware platform, the company announced a lot of stuff — but what it communicated more effectively than anything else was just how incredibly thirsty — one might even say desperate — Mark Zuckerberg is for his metaverse bet to pay off.
Before I get ahead of myself, let me be clear that I understand all of these prerecorded presentations given by large tech companies are extended advertisements. No one’s disputing that, but Zuck’s overscripted and overproduced dev event keynote today was easily the hardest sell for not just a product or a platform, but the premise upon which it’s based, I can ever recall seeing in a decade in tech.
The presentation basically kicked off with Mark assuring us that VR is bigger than ever, though almost entirely in relative terms. It’d be hard for it not to be, given the pace of its growth to date since its advent (counting either from the days of the first VR headsets or what you might call “the modern era,” when the original Oculus Rift finally made its ways to consumers in 2016. It then went into a series of cherry-picked revenue numbers set up by Facebook CTO Andrew “Boz” Bosworth, a direct appeal to an ecosystem in need of fleshing out. These were mostly individual highlights, however, rather than the cumulative rapidly incrementing ecosystem numbers that Apple used to balloon its own mobile App Store efforts through its early days.
The rest of the presentation was basically a series of hand-wavy “announcements” (many reheated versions of prior ones) that aimed to define use cases and domains in which the metaverse and VR would actually be useful to people. Zuck covered all the old stand-bys — social, gaming, fitness and “future of work.” None looked significantly improved or capable of acting as a turning point in terms of mass adoption, and most had either vague or nonexistent ship dates.
One of the biggest swings involved a partnership with Microsoft, which was jointly announced by Zuckerberg and Satya Nadella. Basically, Mark is so desperate to get people in the metaverse that he’s allied with an erstwhile competitor, in a move reminiscent of when Steve Jobs welcomed Bill Gates via satellite link during the Macworld Boston keynote in 1997.
That worked well, but it’s not yet clear if this will. Nadella himself articulated the main and recurring theme of VR: “it’s early days.” Only around 5,400 people were watching in VR when my colleague Taylor tuned in partway through the presentation , just before Mark switched over to using his new avatar and finally actually presenting in the metaverse itself. The avatar itself was a big improvement from prior iterations; it was a fully animated version of the better-looking version Mark showed off after his much-mocked announcement of Horizon’s expansion to France and Spain.
It looked better, yes, but it sure didn’t look like the future.
