Honored to be a part of the Inrupt mission as profiled by Andrew Sears on All Tech is Human (ATIH) CHANGEMAKERS. Here’s a full text version of my interview:

Davi Ottenheimer is the Vice President of Trust and Digital Ethics at Inrupt, a company striving to restore the power of balance on the web through data decentralization technology. For over 25 years, he has worked to apply security models to preserve human rights and freedoms. He is the co-author of Securing the Virtual Environment: How to Defend the Enterprise Against Attack (2012) and the author of The Realities of Securing Big Data, which is due to release this year. Davi spoke with ATIH about where the web went wrong and how decentralization technology can get things back on track.

ATIH: Tell us about your journey to your present role at Inrupt. How did you first become interested in digital ethics work?

My interest in digital ethics goes back to at least sometime in the early 1980s. The 414s seemed to me a foreshadowing of where the world was headed, at least in terms of society defining shades of lawful and unlawful data access. Their story felt very normal, not at all exceptional, because at that time it was similar to what I was experiencing in school and it was on the cover of big publications like Newsweek.

My family also exposed me very early to authorization concepts in both digital and analog tech. I basically grew up seeing computers as a natural next step like a tractor replaces the ox; no one really would want to be without one. That gave me a fluid understanding of ethics across a wide technology spectrum. For example as a child in very rural Kansas we had only a shared “party line” for telephone; my parents would of course tell me it was wrong to listen to our neighbors’ calls. I was fascinated by it all and by the time I was in college studying philosophy I was running my own copper from taps on building wires to connect dorm rooms, shifting underutilized resources to community service by taking over computer labs and all kinds of typical mischief. At the same time I was playfully exploring, I also ended up helping investigate or clean up some clever abuses of the lines by others (e.g. toll fraud, illegal reselling).

More to the point in college I always tried to turn in digital versions of assignments including a hypercard stack (precursor to websites) on ethics of linguistic regulation of Internet hate speech. That felt more exceptional, substantially entering digital ethics, because my teachers sometimes bristled at being handed a floppy instead of the usual paper. I was deep in a world at this time many professors had access to yet barely seen. I still figured at that point since I could dive into it anyone could and soon would. It was around 1990 that I excitedly showed a political science professor a 30 second video clip that I had spent 12 hours downloading and reconstituting. I had been studying information warfare and told him dissemination and manipulation was entering a whole new domain with Internet video… he told me “just do your damn homework” (some typical assignment on Middle East peace options) and walked away shaking his head. I felt at that moment I wasn’t giving up or going back, digital ethics had become my thing.

After college I applied to do political research at LSE and they countered with an offer in the history course. I accepted and explored far more historic cases of ethics in intervention (information warfare by Orde Wingate, and power dynamics in taking over large scale systems while not really owning them — 1940 British invasion of Ethiopia). My history advisor was truly amazing. He encouraged me to go professional with technology work and even told me it wouldn’t be a bad idea to pursue as a career.

It was great advice and I went straight into working for a DEC reseller in California pushing decentralization with PCs and TCP/IP. Getting paid to take hardware and software completely apart to fix it was like heaven for me. From those first phases of interest we can fast forward through twenty-five years of hands-on security within many industries around the world of all sizes and shapes. My journey has always been about convincing people from field to board-level that unsafe technology alters power dynamics, and that we protect liberties by bringing safety principles into engineering as well as policy.

A few years ago a very small database company reached out for help fixing their widely publicized product security flaws. Literally millions of people were being harmed, and they told me they weren’t finding people willing or able to help. I agreed to jump into it on the condition they let me drive end-to-end encryption at the field-level into their product as a feature, while I also cleaned up management practices. It was after we released that end-to-end field-level encryption feature, and after I guided them through IPO and massive growth to a much safer and more proper course including external oversight, that Bruce Schneier strongly suggested I consider the new Inrupt mission to bring Solid to life. I was thrilled to be given the opportunity to join such an important and challenging role.

ATIH: Inrupt is advancing the development of Solid, an open source platform designed to remake the web. What’s wrong with the web that we have today?

Solid presents a powerful yet very simple concept to remake the web: your data lives in a pod controlled by you. Any data generated by you or your things (e.g. watch, TV, car, computer, phone, thermometer, thermostat) goes to your pod. You then control access at a meaningful level, where consent has real power. I call it the need for an off button and a reset button for big data. Don’t want your data used anymore by who knows who? You have that choice. Want to be the authoritative source on data about you? Also your choice. If your doctor wants to look at your fitness tracker data, you grant that. When a family wants to share moments in photos, they grant that. Want your machines to talk with each other, not the manufacturer, and only at certain times? Up to you, through your pod controls.

We expect this to evolve with completely distributed models, although sounding idealistic, because they are necessary and thus not out of the question. At the same time, efficiencies of scale and basic economics tell us many people will have pod service providers instead of going with homegrown or localized varieties. As a long-time self-repair and build-your-own-kernel linux advocate I see no conflict innovating towards both off-grid piece-meal installations, as well as abstract and monolithic cloud services. You win because you have a lot more flexibility in a world where you seamlessly can flow between different worlds of control that suit you.

Sir Tim Berners-Lee calls the Solid project of decentralization being pro-human, as opposed to what he calls the current anti-human web platforms. For me perhaps the best way to explain the current problem with the web might be aggressive centralization, which historians used to say about the neo-absolutist surveillance state of 1850s Austria. I find it useful to reference history events to explain socio-economics that we see today with Google.

Another aspect of the problem, which I have been giving presentations about recently, is how our “digital bodies” owned by large proprietary platforms becomes a form of human trafficking.

Unfortunately 1850s American slave plantations seem to be an appropriate history reference as well. It actually explains Facebook data management and expansionist habits. I don’t say this lightly, especially as Memorial Day was created in 1868 specifically to honor those who made the ultimate sacrifice to win the war that was supposed to end slavery in America.

In my presentations on big data security, for example, I literally ask people to consider how the cotton gin was invented by a woman to end slavery, yet instead it led to state sanctioned rape of American women and forced births to rapidly expand human trafficking. That’s not the kind of history anyone really hears in school, and those are the actual facts. The web was invented to bring freedom, to end our digital selves being locked away, yet it has led to state sanctioned collection methods with vastly expanded proprietary control over almost our entire lives.

ATIH: How did these problems with the web come about?

That’s a great question. There has been massive pressure for data centralization from so many camps that have failed, it’s almost a wonder at all that some succeeded in cornering the web. I’d like to think the problems are the exception (e.g. like nationalization of the telephone under President Woodrow Wilson, or invasive inspection and destruction of U.S. mail under President Andrew Jackson) and we’re course-correcting to get back on track.

Cable television and AOL dial-up services both, believe it or not, were considered threats at some point to the success of a decentralized web. Microsoft too, although it obviously found itself in US government regulatory trouble when it aggressively tried to center the web around its browser and operating system. Some might point to RFC2109 but I find the socio-economics to be more important than this technical specification that helped build statefulness.

Perhaps the real turning point that set back decentralization came soon after the web was being panned as just a fad that would never rebound after the dot-com disaster. We witnessed in a time of crisis the giant transfer from small businesses to conglomerates, which might feel familiar to economists looking at today’s pandemic.

The optimism of the hugely diverse commercialization efforts by startups, which in a large part led to the crash, generated a kind of popular herd momentum that was picked up by the few dominant remaining technology firms. They in fact roared out of the dot-com crash with far more influence, far more human connectivity, and the market rewarded a kind of fast monopolistic growth as it escaped financial downturn. The web’s standardization and ease of use, once transformation to it was popular, made it a perfect vehicle for massive-scale growth.

The next market crash, from the mortgage crisis, then served as another accelerator on the trend for centralization coupled with more powerful devices becoming less expensive and default connected to the standards-based web. The technology sector became seen as a stable financial engine and attracted business innovators who believed user generated content had the best potential value and they set out to build systems that flipped the web on its head; would keep users connected by making it difficult to exit.

What’s notable about this history is the financial conditions and technological shifts that may never again materialize in quite the same way. That’s why I see dangerous centralization as a form of regression, an error that requires applied humanitarian correction. It’s like firing a CISO who steals, or countering the rise of extremist anti-science cults that typically form in response to recent scientific breakthroughs. I don’t believe in an inherent centralization need, or natural monopoly, in this context. In fact I see more the opposite, that we should think about Facebook in terms of why abolition of slavery never should even have been a debate. Had there not been the stress effects that led to over-centralization as a form of wealth preservation (arguably an incorrect response fueled by other unfortunate market conditions) the web could have continued to evolve in the more naturally pro-human model.

ATIH: Inrupt’s Co-Founder, Sir Tim Berners-Lee, calls Solid a project “to restore the power and agency of individuals on the web.” How does Solid accomplish this?

Giving users consent controls over their data, making the technology that represents a human to be actually human-centric, is the path forward. The pod is a representation of one’s self that should be liberating, and is manifestly difficult to do without Solid. Many people now seem to agree we can and need to fundamentally alter the balance of power, whether they call it Solid or something else. Given that everywhere you are something is a computer generating data that represents you, taking control of your digital body has become essential.

In terms of projects at massive scale I’ve lived through early days on multiple data sharing standard initiatives (DICOM, TCP/IP, HTTP, HTML, KMIP) and worked deep inside the security teams for the biggest data platform companies (EMC, Yahoo, ArcSight, etc) so much of the scope and ambition feels familiar. I mean when you really think about it, the Solid protocol could be even bigger and more exciting, like a 1968 Carterfone moment or even 1862 Emancipation Proclamation, in terms of how small decisions restore power to individuals at massive scale.

The market either will keep building costly bilateral bridges and proprietary hubs people are trapped on, like we’re seeing with Apple and Google attempts at a Covid-19 alerting API. Or standard languages and protocols will emerge to lower barriers to human-centric innovation and expand the web again by decentralizing it and shifting to Solid. If it’s the latter we’ll see a boom in more generalized global prosperity. Trust is an essential step towards making it all work, just like always, yet user security awareness levels are so much higher than I’ve ever seen before. Doing things the right way for humans to preserve their agency, to help everyone avoid becoming victims of digital trafficking, brings up all kinds of granular authorization discussions. We are quite literally reimagining technology in ways that the Solid protocols will better augment and protect positive human conditions, complex communities, cultures and all.

ATIH: What does your role as VP of Trust and Digital Ethics look like day-to-day?

I get up and wonder if I can configure my toothbrush to start streaming its data to my pod. Pour a cup of tea and think about a digital assistant reading data in my pod to know exactly what leaf and temperature “Earl Grey, Hot” means, and how accurate automated speech recognition would become when designed from the start to be highly localized on pod data and trained for my particular accent within my community or even my family.

But seriously, every day is running a gamut of working on trust concepts for the specification and protocols being designed, including security for the related products being engineered and the services offered to support both at a global production-ready scale.

People are coming to us with very real problems to solve across every sector and industry. It’s a great feeling to be helping them, working with such a passionate team that carries a real vision of a better future. Lately there has been a lot of discussion with both large and small organizations that have similar values and find Solid and Inrupt deliver a valuable piece in their own human-centered technology development projects.

Almost every day I am reminded of the important lessons and legacy of John James Ingalls such as his application of “ad astra, per aspera”, roughly translated as purposefully shooting for the stars, which became the Kansas state motto in 1861.

Fine issued for misuse of fingerprints.

The logic of this huge enforcement action was simple, biometric data was collected disproportionate to need.

Employees of a company had to have their fingerprints scanned for attendance and time registration. After investigation, the Dutch Data Protection Authority concluded that the company should not have processed employee fingerprints. The company cannot rely on an exception ground for the processing of special personal data. The company will be fined 725,000 euros for this.

Humans were at put risk because privacy wasn’t being properly minded. Attendance and time authentication were not reasonable use-cases, as they have effective ID options that do not need collection of biometrics.

Exception for collection would be made if fingerprints were an appropriate control mechanism, such as in a system protecting the user’s data by verifying them by something they are.

The numbers are expected to go even higher, but for now the NYT has said it’s reasonable to assume 20,000 people in NYC were killed in just 30 days.

Part of the reason for the revised count has been COVID-19 visualizations that compare current death rates against historic ones.

The empire state building looking thing on the right is the rate of death in NYC during COVID-19 relative to the historic rates to the left.

The Financial Times did a similar analysis globally.

Global coronavirus death toll could be 60% higher than reported: Mortality statistics show 122,000 deaths in excess of normal levels across 14 countries analysed by the FT

This kind of comparison of current deaths against historic averages seems an extremely wise way to estimate severity of COVID-19 right now for several reasons:

First: COVID-19 autopsies have confirmed the early signs that the virus kills people in novel ways.

We already heard that EMT crews couldn’t keep defibrillator batteries charged during a single shift because cardiac arrest calls during the pandemic suddenly tripled or higher.

EMT also reported unexpected trauma of low survival rates.

The cardiac arrests are the hardest calls right now. More than once, we have been present at the moment of capture and yet were unable to save the patient. In the past, if a patient goes into cardiac arrest and we witness it or are there within three minutes, we can often save them. We use a defibrillator to shock them and restart their hearts. But for COVID-19 patients, this is not happening. We are not getting any of them back — and now the Department of Health doesn’t want us to bring dead patients to the hospital, so we are pronouncing them dead in the field and turning the bodies over to the police who have to wait for a coroner.

Second: CDC has started to release reports that even early February deaths in California homes were from COVID-19

Officials say they originally thought that the first COVID-19 death in the [Santa Clara] county was on March 9. Autopsies were performed on two people who died on February 6 and February 17. The CDC received tissue samples from the coroner and were able to confirm that both cases were positive for SARS-CoV-2. The third individual who died on March 6 was also confirmed to have been positive for COVID-19.

That death at the start of February was a woman who had a “burst heart”, further proving the point above about novel ways of COVID-19 killing people.

County health officials have said if they knew at the time the woman had coronavirus, they might have issued shelter in place orders earlier. […] “There’s an indication the heart was weakened.” [Dr. Judy Melinek, a Bay Area forensic pathologist who reviewed the autopsy report] said “The immune system was attacking the virus and in attacking the virus it damaged the heart and then the heart basically burst.” Dowd’s husband, citing his wife’s strong exercise habits and overall good health before falling ill, had requested an autopsy.

NYC still maintains that March 11 was their first date of death for a confirmed death, which obviously will need to be changed.

Given how a healthy American abruptly died February 6th from the virus, consider also how the White House was operating at that time. Just Security provides a detailed timeline:

February 10-March 2, 2020 … five rallies across the United States, each attracting thousands of attendees in confined spaces. The rallies take place in New Hampshire (2/10), Arizona (2/19), Colorado (2/20), Nevada (2/21), South Carolina (2/28), and North Carolina (3/2).

These rallies, like a death cult gathering, will most certainly be a cause of fatalities in America.

Third: Low numbers are controversial. I’ve been tracking death rates from the NYC Department of Health since the first cases reported (tragically unreported in the JHU dashboard, as I wrote about March 3rd).

When I posted this following chart the other day, for example, I immediately heard backlash from people with family in NYC. They complained deaths were known to be very high so there was no possible way my graph could have such low numbers showing a decline let alone tapering off.

Red is death, Grey is hospitalization.

It’s true, while the actual death rate is high, it likely is even higher than what these official NYC Department of Health numbers show. Confirmed COVID-19 test results increasingly looks like a subset of deaths far above normal trends compared to death rates of prior years.

I’m not saying the low count graph I made is wrong in terms of a trend. That trend is real and does reflect the case load on NYC services. The numbers definitely are in decline and pressure is considerably lower on EMT.

What’s surely low confidence is the daily count. When I can find the data and the time, I will add in a low/high estimate to show actual deaths daily and not just the shape of the pandemic curve.

One final thought. Often when I post a visualization of deaths some middle-aged white man invariably will come forward and say a per capita rate is the only thing that matters. Imagine a close relative dying and some random guy says to you “don’t worry, your sister’s death per capita is insignificant, given how siblings overall in this region are doing just fine”

Having to care about others drives some people to minimize human life through “per capita” models. Leaving off the per capita calculation tends to reveal callous and selfish thinking by viewers.

Per capita still has a place. Experts are good at finding ways to make different population numbers relevant (to measure likelihood or severity) yet that shouldn’t be turned by just anyone into a license to dismiss every human life as a percentages game.

The better model is vision zero, which says 40,000 American traffic deaths per year is 40,000 too many.

In November 2019 a DIY article was posted on how to build and train an inexpensive RPi thermal camera.

Even with more complex network architectures, the optical model wouldn’t score above a 91% accuracy in detecting the presence of people, while the thermal model would achieve around 99% accuracy within a single training phase of a simpler neural network.

Despite the high potential, there’s not much out there in the market — there’s been some research work on the topic (if you google “people detection thermal camera” you’ll mostly find research papers) and a few high-end and expensive products for professional surveillance. In lack of ready-to-go solutions for my house, I decided to take on my duty and build my own solution — making sure that it can easily be replicated by anyone.

Now you can easily make one to mount on your door and give thermal readings for guests as well as announce known visitors.

The commercial FLIR thermal camera site gives this image as proof of its utility, although I expect they soon will update to reflect pandemic uses as well.

Not surprisingly, despite the long list of reasons to use thermal imaging (e.g. higher integrity of signal, more resilient to environmental interference) the EFF makes a very tone-deaf argument against its future use:

Terrorism is one thing — because it’s an ongoing problem. But there’s no reason why this kind of technology would need to stick around after the COVID-19 crisis is over.

That reads to me that the EFF believes after COVID-19 crisis is over there will no longer be any other threats, let alone a need for higher integrity in visual signals (e.g. authentication).

In a breathtaking move of transparency, the White House has come forward to reveal the head of its COVID-19 policy and response coordination all along.

Behind the scenes, held as a tight secret until now, was the highly decorated and very well known General Buck Turgidson. The General formerly had led efforts to drive the world towards global cyber annihilation.

The American government plan to delay its COVID-19 response is said now to have been intentional, pushing to the highest death rates in the world within just one month.

Mass casualties was estimated to have the effect of positioning US government as the most helpless victim, to set up vicious attack campaigns and fire angry missives against China and WHO.

Proud of delays and confusion of the American people, that led so quickly to tens of thousands killed, the White House posted also the following example of detailed analysis from the General:

Mr. President. I’m not saying we wouldn’t get our hair mussed. But I do say no more than 10 to 20 million killed, tops.