Tech Brew Ride Home - Fri. 08/06 – The Apple Photo Scanning Controversy

Episode Date: August 6, 2021

I try to explain this whole controversy about Apple scanning your photos. Yelp will list business’ vaccination policies. I guess I’m bullish on autonomous tractors. And, of course, the weekend longreads suggestions. Sponsors: Quantummetric.com/podoffer offer code PODCASTCODE Metalab.com Links: Apple confirms it will begin scanning iCloud Photos for child abuse images (TechCrunch) Yelp will let businesses list their vaccination policies (The Verge) John Deere buys autonomous tractor startup Bear Flag Robotics (TechCrunch) Weekend Longreads Suggestions: The End of Venture Capital as We Know It (The Information) Deal Of The Century: How Michael Dell Turned His Declining PC Business Into A $40 Billion Windfall (Forbes) The Metaverse Has Always Been a Dystopian Idea (Vice) How Olympic Surfing Is Trying to Ride the Machine Learning Wave (WSJ) Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to the Tech meme right home for Friday, August 6th, 2021. I'm Brian McCullough today. I try to explain this whole controversy around Apple scanning your photos. Yelp will list businesses vaccination policies. I guess I'm bullish on autonomous tractors and, of course, the weekend long read suggestions. Here's what you missed today in the world of tech. Apple says in the next couple of months, it will begin scanning photos on iOS 15 and macOS Monterey devices in the U.S. for, known child abuse images. Quoting TechCrunch. Apple told TechCrunch that the detection of child sexual abuse material, also known as CSAM, is one of several new features aimed at better protecting

the children who use its services from online harm, including filters to block potentially sexually explicit photos sent and received through a child's eye message account. Another feature will intervene when a user tries to search for CSAM-related terms through Siri and search. Most cloud services, Dropbox, Google, and Microsoft, to name a few, already scan user files for content that might violate their terms of service or be potentially illegal, like CSAM. But Apple has long resisted scanning users' files in the cloud by giving users the option to encrypt their data before it ever reaches Apple's iCloud servers. Apple said its new CSAM detection technology neural hash instead works on a user's device and can

identify if a user uploads known child abuse imagery to iCloud without decrypting the images until a threshold is met and a sequence of checks to verify the content are cleared. Apple is trying to calm privacy fears by baking in privacy through multiple layers of encryption, fashioned in a way that requires multiple steps before it ever makes it into the hands of Apple's final manual review. Neural hash will land on iOS 15 and MacOS Monterey slated to be released in the next month or two and works by converting the photos on a user's iPhone or Mac into a unique string of letters and numbers known as a hash. Anytime you modify an image slightly, it changes the hash and can prevent matching.

Apple says neural hash tries to ensure that identical and visually similar images, such as cropped or edited images, result in the same hash. Before an image is uploaded to ICloud photos, those hashes are matched on the device against a database of known hashes of child abuse imagery provided by child protection organizations like the National Center for Missing and Exploited Children and Others. Neural hash uses a cryptographic technique called Private Set intersection to detect a hash match without revealing what the image is or alerting the user. The results are uploaded to Apple but cannot be read on their own.

Apple uses another cryptographic principle called Threshold Secret Sharing that allows it only to decrypt the contents if a user crosses a threshold of known child abuse imagery in their iCloud photos. Apple would not say what that threshold was, but said, for example, that if a secret is split into a thousand pieces and the threshold is 10 images of child abuse content, the secret can be reconstructed from any of those 10 images. It's at that point that Apple can decrypt the matching images, manually verify the contents, disable a user's account, and report the imagery to NCMEC, which is then passed to law enforcement. Apple says this process is more privacy mindful than scanning files in the cloud as neural hash only searches for known and not new child abuse imagery.

Apple said that there is a one and a trillion chance of a false positive, but there is an appeals process in place in the event an account is mistakenly flagged. Apple has published technical details on its website about how neural hash works, which was reviewed by cryptography experts and praised by child protection organizations, end quote. In addition to scanning for known child abuse photos, Apple will apparently also begin using on-device machine learning to warn parents and kids of sexually explicit photos in iMessages. So obviously, this is all good, you know, protecting children, no one's against that. But to say that some security experts are concerned that this might have secondary effects to privacy is, well, kind of putting it mildly. Many privacy experts are saying governments will only increase the scope of this type of scanning,

which is essentially a government accessible backdoor, right? Which isn't that what Apple has fought against for so long? Doesn't this feel like a wild about face for them? Let me read you a thread from Danny O'Brien, who is a special advisor to the Electronic Frontier Foundation. So for the last few years, intelligence agency experts have been pushing for tech companies to put remotely controlled software to scan for unlawful content in people's devices rather than surveil on the network. The reason is that government departments, including law enforcement and the intelligence community, want to be able to peer into all encrypted material

as easily as they used to, briefly, spy on web and email traffic back in the, quote, golden age of mass surveillance, which ranged roughly from 9-11 to 2013. The thinking goes like this. Modern encryption works end to end, from the sender to the recipient, or at, rest, encrypted only on your device, with only you having the key. So why don't we put the government intercept point on one or both of these ends? The euphemism for this is client-side scanning. But what it requires is code to run on your own device that you can't turn off or control, because if you could, then of course that's the first thing the bad guys would do. Companies, cryptographers, and human rights groups have pushed back hard against this.

Cryptographers because it's basically a backdoor, human rights groups because of the huge abuse potential and weakening of privacy norms and companies because they depend on users' trust. That's why some countries, the UK and Australia by law, China and others through state pressure, want to force client-side scanning on companies by law. To my knowledge, no one so far has actually done this because they figure the outrage and pushback would be too extreme. Now, without an explicit order from any government that we know of, Apple appears to be putting client-side scanning into their device software. I don't know the details, but I imagine Apple has made this as limited as they felt they could do, but a line has been crossed. I'm worried for how much this means I can trust Apple in

the future, especially my Hong Kong friends who are already dealing with Apple's attitude to their secure communications. But I'm far more concerned about the pressure this puts on other hardware and software vendors. I really wonder how Apple is going to message this or whether there's something I'm missing. It seems to go against their stance in the FBI San Bernardino, case, their claim to be a safe custodian of their user's privacy, end quote. Yelp is going to let businesses list their vaccination policies right inside Yelp's search results, including if customers must show proof of vaccination, wear masks, or if all staff are vaccinated, quoting the verge. Businesses can list the attributes via their Yelp for business

accounts, allowing users of the service to find places with policies they're most comfortable with. The ability to list vaccination policies joins a long list of pandemic-focused features. Yelp has added to its service over the past year. Early on in the pandemic, the service let businesses specify whether they offer virtual or contact-free services and later expanded this to other safety measures, like whether they offer outdoor dining or have a face mask policy for staff. Vaccination policies have become controversial in some circles, so Yelp is also emphasizing the moderation work it does to prevent businesses from being review-bombed as a result of their

choices. It says it's proactively monitoring the pages of businesses that choose to display these policies and that it will remove reviews that focus on them rather than a customer's actual experience with a business, end quote. Interesting raise time. I don't know how bullish I am about robots in general, and if you listen long enough, then you know that I'm skeptical of self-driving cars and claims that autonomy is right around the corner, but somehow I'm super bullish on robots and autonomy getting big in construction and agricultural settings. Because, you know, you don't have to worry about running over people or running into things in the middle of a field somewhere. So agricultural manufacturing giant John Deere acquiring autonomous

tractor startup bare flag robotics for $250 million, caught my eye, quoting TechCrunch. The Bay Area-based firm, which specializes in autonomous farming heavy machinery, was founded in 2017. They first crossed our radar the following year as a member of Wycombinator's winter 2018 cohort. We got a tour of an orchard and just how pronounced the labor problem is, co-founder Aubrey Donnellan told TechCrunch at the time, they're struggling to fill seats on tractors. We talked to other growers in California. We kept hearing the same thing over and over. Labor is one of the most significant pain points. It's really hard to find quality labor. The workforce is aging out. They're leaving the country and going into other industries, end quote.

In the intervening years, John Deere tapped Bear Flag for its own startup collaborator initiative, and the robotics firm has also begun to deploy its technology in an undisclosed, limited, per their wording, number of sites in the U.S. Agricultural is one of several robotics categories that have seen a spike in interest in the past year due to labor shortages that predate but were exacerbated by the global pandemic. Of course, that interest doesn't make anyone immune from the difficulties of launching a robotic startup. Last month, Apple picking robotics firm, Abundant, confirmed it was closing up shop, noting, quote, after a series of promising commercial trials with prototype Apple harvesters,

the company was unable to raise enough investment funding to continue development and launch a production system, the company noted at the time. An acquisition seems like a reasonable outcome for a company like Bear Flag. The startup gains a lot of resources from its massive new owner, and its new owner adds some new tech to its portfolio. Indeed, John Deere has been pretty aggressively looking to expand into more cutting edge technologies like robotics and drones in recent years, end quote. Time for the weekend long read suggestions. First up, this is another instance of a story behind a hard

paywall. But if you care about venture capital at all, this piece from Sam Lesson, the end of venture capital as we know it, is the title, is all anyone that is in the venture capital industry has been talking about all week. Quote, what is really happening is, that capitalism is functioning as intended, and it has worked throughout history. The era of West Coast style venture capital that has been shaped by the growth of the software industry is coming to an end. In any new market, venture capital has come in and provide very expensive capital for high-risk, high-reward propositions at the frontier. Over time, these investments in new industries become better understood and instrumented. Their risks and opportunities can be more easily

measured, and investors across the board price them more or less the same way, as these shifts occur, massive flows of capital follow, and investors compete with each other to offer industry builders cheaper and cheaper money. This happened in the New England whaling industry in the 19th century in what was arguably the first VC cycle, and it is happening today in software. What we are currently calling venture capital is rapidly becoming just run-of-the-mill, globalized, highly competitive, and reasonably low-margin finance, end quote. Next, I'm old enough to remember when Dell was the stock, every fund, manager had to own because it had this funny habit of going up year after year after year,

at least in the early 90s. But then Michael Dell himself took Dell private and what was once a key player in the tech industry seemed to disappear from relevance. And yet, as this big profile in Forbes shows, that is probably completely wrong. Dell might represent the biggest private equity success story of all time. Quote, Michael is financially sufficient. He's not a technology geek by any stretch of the imagination, says George Roberts, the billionaire co-founder of private equity giant KKR, and a pioneer of the leveraged buyout who marvels at the deal. Quote, he bought the company back at the right time. With hindsight, his timing looks pretty perfect to me, end quote. At 56, Dell is technology's last man standing, the final original founder of

the computer era still running his baby. His rivals have aged out or moved on, whether tech billionaires Bill Gates or Larry Ellison or Steve Balmer, who have shifted course to philanthropy or trophy assets such as Hawaiian Islands and NBA teams. Everybody's eyes are on Amazon, Microsoft, and Google says billionaire Mark Benioff, the co-founder of Salesforce and a friend of Dell's. They don't realize that Dell has quietly amassed the market share in enterprise technology, end quote. Then, given our discussions of the Metaverse this week, I thought this piece from Brian Merchant in Vice was provocative. The Metaverse has always been a dystopian idea, Brian says. Quote, In the world of Snow Crash, the Metaverse is not viewed as particularly cool. It is necessary

because the real world has become so unbearable. Ditto in the most famous book to update the metaverse's architecture for our modern pop culture saturated era, Ready Player 1. Its Oasis is basically of the Metaverse if it were written by a neural net trained on 80s movies and OTS-era video games. Both books Metaverses get at a common truism. There is something inherently dystopian in a future where humans abandon the real world in favor of an escapist and consumers-oriented, fully immersive digital one. To want to spend any serious amount of time in a metaverse, it must be made more appealing than reality, a feat which can be accomplished in one of two ways. Either the world outside is already shitty enough to drive you into

a glitch-prone, murder-filled alternative, or the fantasy of becoming someone else is compelling enough to consume you totally, end quote. And finally, how machine learning and artificial intelligence is transforming the new Olympic sport of surfing, quote, wave forecasting is among the biggest applications machine learning has had in surfing to date. The use of wave Forecasting technology has a long history that precedes surfing, including in coastal engineering, shore protection, and combat planning for World War II. According to Westwick, the historian, wave forecasting has more recently benefited from some of the same technological advances that have enabled other commercial machine learning technologies

like image and voice recognition, namely the large-scale availability of data and computing power says Ning Li, an Ocean Wave model's system specialist at the University of Hawaii at Manoa's Pacific Islands Ocean Observing System, end quote. All right, so no Ride Home Plus content this weekend, although we do have an interesting Rayses episode in the pipeline for next weekend. Everybody is going to get access to that Twitter space that Chris and I did on Wednesday, deep diving into the Metaverse. Other than that, enjoy your weekend.

Talk to you on Monday.