Second Sight left users of its retinal implants in the dark | Hacker News Reader

Second Sight left users of its retinal implants in the dark

spectrum.ieee.org

276 pointsheadalgorithm85 comments

I've said before that I think it should be illegal to sunset a device without open-sourcing the technology needed to support it. Medical implants of any kind are an especially compelling case for this

Probably medical devices should also have very strict requirements regarding auto-updates, including provisions that prevent them from being remotely disabled

This seems like a problem we can't solve without legislation. We need either these sorts of legal requirements for implantable devices, or to blanket ban private firms from working in this space. I think the former would be preferable

> including provisions that prevent them from being remotely disabled

From the article, none of these implants were remotely disabled. The article's first paragraph is deceptively written to make it seem like that though. Further down it clarifies:

> But in 2013, after four years of regular use, Campbell’s system shut down in the subway station, and despite some repair attempts by Second Sight, never worked again.

So it sounds like a random hardware failure, which the company attempted to repair, but failed. Also in 2013 Second Sight was still 6+ years away from going bankrupt. Seems like this implant wasn't replaced/repaired because it would have required another surgery, not because of the bankruptcy.

The article is incredibly misleading. They don't mention the year that the events in the first paragraph happened until halfway down the page and even if you read that far you might not have connected the two.

Her implant failing is not even related to the company going bankrupt (although maybe it was a precursor)

The article is not even a slightly bit misleading, if you read it in full and without any prior expectation or reading other comments. Here is a quote: "took hundreds of patients on a roller-coaster ride of technological innovations, regulatory successes, medical and financial setbacks, and a near-total meltdown" – this is not simply about how they abandoned the tech or almost went bankrupt. This is also about the challenges they and their patients faced along the way.

As other comments have mentioned, the article is misleading. No software or hardware was remotely disabled. Parts failed over time. Once the company went out of business, support was no longer available. Yes, this is terrible, but there was no malicious intent. If anything they tried mightily to support their customer base despite losing money the whole time.

More importantly: If you require that medical devices be supported indefinitely, you will heavily discourage anyone from making medical devices. Such a requirement is an endorsement of the Copenhagen interpretation of ethics.[1] Had this company not existed, these people would have been blind for the rest of their lives. Instead, this company tried to help people and ran out of money. The net effect is that blind people were able to see for years. This is much better than the alternative.

If you want to assign blame, assign it to the FDA. They're the ones who make it practically impossible to roll out any new medical technology. Their byzantine rules prevent the sale of devices as simple as epipens.[2] Their requirements (though well-intended) drastically increase treatment costs and slow the pace of progress.

1. From https://blog.jaibot.com/the-copenhagen-interpretation-of-eth...

> The Copenhagen Interpretation of Ethics says that when you observe or interact with a problem in any way, you can be blamed for it. At the very least, you are to blame for not doing more. Even if you don’t make the problem worse, even if you make it slightly better, the ethical burden of the problem falls on you as soon as you observe it. In particular, if you interact with a problem and benefit from it, you are a complete monster. I don’t subscribe to this school of thought, but it seems pretty popular.

2. https://slatestarcodex.com/2016/08/29/reverse-voxsplaining-d...

> As other comments have mentioned, the article is misleading. No software or hardware was remotely disabled.

Nothing in the article is misleading in this regard, it says nothing about remote disabling of the systems. This is entirely a creation of this thread, and not the article. The word "remote" doesn't even show up in the article. Your comment is the one that's misleading by asserting things about the article that aren't even remotely true.

My reading of your parent comment was that "no software or hardware was remotely disabled" was intended to mean, "there was nothing even a little bit like dialing hardware or software," not,"no hardware or software was disabled from a distance."

> My reading of your parent comment was that "no software or hardware was remotely disabled" was intended to mean, "there was nothing even a little bit like dialing hardware or software," not,"no hardware or software was disabled from a distance."

How is that a reasonable reading of their comment?

They wrote:

> No software or hardware was remotely disabled.

How can that not mean:

> "no hardware or software was disabled from a distance"

"from a distance" and "remote" are, pretty much, the same thing in the context of computer systems.

Because there's another meaning of "remotely," as in the phrase "that's not even remotely true." Not having anything to do with physical distance, and referring to metaphorical distance from the facts.

but that is not remotely how the phrase remotely disabled is ever used.

> More importantly: If you require that medical devices be supported indefinitely, you will heavily discourage anyone from making medical devices. Such a requirement is an endorsement of the Copenhagen interpretation of ethics.[1] Had this company not existed, these people would have been blind for the rest of their lives. Instead, this company tried to help people and ran out of money. The net effect is that blind people were able to see for years. This is much better than the alternative.

I absolutely agree, which is why that is not even remotely what I proposed. Rather than require indefinite support, we should require that support be made possible by the owners of devices, or, when this is too technically demanding, anyone with the know-how and desire to help (out of kindness, profit motive, whatever). Currently, this is very difficult because devices tend to require proprietary information to maintain (such as software). Thus, I believe it should be a legal requirement to open-source and make publicly available the materials needed to maintain/modify/repair the device (e.g. firmware, internal tools) once you've stopped supporting it. Right now, maintaining an unsupported device requires serious reverse-engineering skills, and usually plying those skills for that purpose is illegal under the DMCA's anti-subversion rules. I believe such a measure to be necessary to reduce the insane amounts of waste companies create by releasing proprietary devices and then obsoleting them so that they will inevitably stop working, but the mere existence of cases where a device is critical to some function (like a medical implant) constitutes another compelling reason why this requirement is necessary, even if this particular article about this particular device is somehow misleading

I will add (because it's fun to be provocative on the internet) that this is, I think, a reasonable compromise from the position I hold emotionally (which is that all proprietary software is a user-hostile scam and should be outlawed in its entirety) which still accomplishes not leaving devices that have lost support inoperable junk

EDIT: I'd like to add that it's super uninteresting to me to talk about "malicious intent" with respect to corporate actors. They're not people, they have no capacity for malice. I don't believe such an entity's intentions ever matter. They are machines built out of rules and only ever are guaranteed to serve the purpose of trying to make someone money. This is fine, but it's important to ensure that they are built safely, as with any machine

How on earth do you mismanage a company so badly that you can literally give sight to the blind and still go bankrupt? Even at a $500k treatment there should be more than enough aging millionaires willing to buy their way out of their own body's infallibilities

this doesn't surprise me at all. I've had many negative experiences with companies that make disability aids. the target market is poor and vulnerable, they're often monopolies, and they can charge tons of money for crappy tech through insurance.

I bet insurance companies declined to pay what they wanted and they decided to pack up shop in search of a bigger exit. wonder if they ever lose sleep about their abandoned users.

All this, plus the fact that there are lots of pathologies which can lead to similar results, so the effective target market for assistive devices tends to be much smaller than even the small market originally expected.

Further shrunk by communities “closing up” from a combination of having been burned too many times, discrimination / in-group assistance, and viewing attempts to “fix” them as ableism, accommodation refusal, and further efforts to shun them / discriminate against them / divide them. See e.g. colchear implants in the deaf community.

They can't really give sight to the blind - from the article: Normal vision, this is not. Patients and doctors alike stress that the Argus II provides a kind of artificial vision, really a brand-new sense that people must learn how to use. Argus II users perceive shades of gray that appear and disappear as they move their heads.

So it's not like millionaires with blurring are going to be getting in line. This seems like it's for the totally blind.

Perceive also:

https://secondsight.com/discover-argus/

https://www.marketwatch.com/investing/stock/eyes/financials?...

Probably because they weren't the only company working on this and selling products and even at $500K they weren't making money:

it faced financial headwinds. Second Sight was selling the Argus II for around $150,000 in the United States—about five times as much as other neuromodulation devices, according to Greenberg. But even so, he says, the company was losing money: “With all the overhead of sales and regulatory people, it wasn’t profitable.”

Fascinating article, definitely prompts thoughts around the abandonment of technology...

From a technical standpoint, I wonder how hard it would be to reverse engineer the wireless protocol that the implant is expecting, so that a custom camera/software could be used to beam in the data?

That wouldn't help if the implant fails, but for cases like Jeroen (from the article) who dropped his VPU and needed to crowdsource a replacement, it might be a godsend.

Yes, definitely would be worth it to try and dump the code from a broken VPU and do whatever we can to reverse it.

I wonder if a small hacker collective could get in touch with him?

Frankly, I'm not surprised there was no backup plan in case the company went out of business. The quote from the engineer is telling, “We didn’t sell any more, we didn’t make any more, we didn’t have anything to do with it anymore.”

Like a company discontinuing a model of Bluetooth speaker or size of tire, as if the implant isn't a permanent part of their patient's bodies. Just leaving a small division that supported existing users, even with just spare parts, would not have been cost prohibitive.

It sounds like the suits demanded profitability and literally left them customers in the dark. I wonder if in the coming years and more and more biotech implant companies come to market some will be required for long term support. Maybe their relevant IP becomes public domain if the device is not "supported"?

This is a perfect case study on what should be required (legislation??) and considered before licensing a biotech company. These all should have the strictest warranties in the end user's favor.

I know I know, wish in one hand, shit in the other. See which fills up faster.

We have laws concerning doctors, hospitals and codes of medical ethics that govern health care.

I am sure that there are cases where a hospital shutdown due to bankruptcy without transferring their patients but that is not the norm. We should hold other companies in this space to the same standard.

Edit: >To be maintained to satisfy retention requirements under federal and state laws. Maintenance of medical records is especially significant if the organization is defined as a Covered Entity 1 under HIPAA. Failure to maintain and protect medical records can lead to significant fines and penalties even during organization closings and bankruptcy proceedings.

http://metalquest.com/files/Medical_Record_Considerations_Du...

I don't think companies should be allowed to sell products like this (requires invasive surgery, long term maintenance, more invasive surgery if something goes wrong) directly to patients.

They need to sell to the government/health insurance companies, and provide all the tools/training/equipment for those organizations to do the long term support for patients.

Honestly this seems like a good idea. If a company stops selling/supporting a product (or never starts), its IP rights should be forfeit. Require them to be sold to a company willing to sell/support the device, or they should become public domain. This change would probably fix most issues with current US IP law.

How would that keep the company in business to provide support? Seems like the same outcome, to me.

They should be forced to fund a trust that can support all users for their lifetime.

How much do you reckon that would cost per user?

A fairly large amount, even for minimal support. If it's prohibitive to create cool medical devices due to this, the government can assist (guaranteed loans against future profit? whatever, I'm sure there's a bunch of ways to structure it).

Maybe it will push these companies to make support easier as well, and Open Source parts of it where it makes sense to.

I am torn. On one hand, this is great idea (exactly for such cases). I feel unbelievably bad for users.

On another hand. Lifetime support of technology (say 50 years) is damn expensive. Let say you need a dozen of engineers (hardware + software) + doctors to keep it going. It's 12 people * 200k salary (if you don't like this number, pick your own) * 50 years = $120M.

I think you can potentially argue that you don't need a dozen people to support this. However, I think it's a fair number to support aging hardware + software.

Unfortunately, such things work only with a scale. You need 10 people to support 100 customers, 20 people to support 1000 customers and 40 people to support 10000.

If they got to 10000 customers, I could see them getting enough money to fund such trust. Having just 300 customers won't be enough.

Looking at the crunchbase (https://www.crunchbase.com/organization/second-sight). They whole funding was $130M.

Perhaps this could be accomplished by making them front or insure the cost of the expected liability — being at least the cost of removing unsupported devices.

It wouldn't necessarily, but it'd allow others to fill the niche should the company stop filling it themselves.

My understanding is that copyright law is meant to allow a company to profit off its ideas for a set period of time, in order to recoup R&D costs. Then, if the company no longer has the means to profit off the IP, and won't sell it to another who wishes to make the IP available, the IP should become public domain.

Easy to say but what about paying creditors? The legal system we have now needs to be radically rethought to allow a scenario where a company going bankrupt will be forced to give up its patents and copyright on software to the public, instead of selling them to the highest bidder who will then decide what to do. It's not simple, but we need a serious reform in this direction.

I don't see what would be so different than what we have now? A company in bankruptcy is already forced to sell all IP to satisfy creditors. Now the only additional requirement is they sell to a company planning to make use of the IP. If they can't the IP is worthless, and so should just end up in public domain.

I don't think you're considering how loose the definition of 'use' is in this case. They might intend to 'use' the IP in a way that makes them no more able to help previous customers than a patent lawyer is able to build a robot.

I think you're getting hung up on minutia that could be sorted out. The definition of "use" could be sufficiently limited to allow for the company to either support the previous customers, or lose the IP

But even if the new company plans on using the IP, there's no guarantee that the company buying the IP will use it for the same purpose or want to continue supporting the existing use. Like maybe the same technology is useful for implants to provide artificial hearing to deaf people and the new company wants to use it for that.

So if the sale of IP is allowed because the new company will use it, that still leaves the visual implant users in the dark (no pun intended), but if the sale of IP is not allowed or is placed into the public domain to allow visual implant users to use it, then the original company was deprived of the value of the IP.

Presumably the company that buys the IP would also be obligated to support the patients. And would have to prove their capability before the sale would be allowed.

And if nobody qualifies, then even if there are willing buyers the IP would be public domain'ed.

This is likely to have the unintended side effect of reducing research -- investors will be less likely to put money into a company if they'll have severe restrictions on what they can do with the IP if the company doesn't succeed and they may even have there millions of dollars of investment released to the public domain because they couldn't find a company willing to continue with an unsuccessful use of the IP even if it would be valuable for other uses.

The restriction is really “use it or lose it”. I don’t see how that would influence anything. If they are investing and plan to use it, then they retain all value. If they chose not to use it, they can sell it to someone who will or release it into public domain.

There is still a huge potential for earning by investing in something like this, even with restrictions as described.

If they chose not to use it, they can sell it to someone

Well, they can sell it to someone who is willing to continue supporting it for the use that the original company couldn't make work.

or release it into public domain

Yeah, that "lose it" part is the problem - if they can't find a company that's willing to continue supporting what was already proven to be an unsuccessful use case, then their IP has no value.

It is lose it. That’s the point.

> If they chose not to use it, they can sell it to someone who will

You're assuming that someone else can make it work financially in the original domain, but what if no one can?

The IP may have value in other domains, but you seem to be saying that it goes into the public domain if no one can make it work in the original domain.

If that's the rule on IP with medical uses, medical applications just went to the bottom of the list of uses. In fact, you'd be stupid to go for a medical application of IP that is used elsewhere.

The problem is that the device and support associated with it outrageously expensive. Even if the company want to support their users, it may not be sustainable.

This sort of thing required ongoing government support to sustain until the technology is ready.

It seems more like proof of concept device that belongs in a lab rather than a commercially available technology.

"cute" titles should be reserved for tabloids. Say the real thing as best you can in the title, especially when human (or any creature) suffering is involved.

Four reactions: 1. I feel bad for those who had the chance to see and then lost it.

2. Sounds like a normal bankruptcy, where all the focus is on making sure someone gets their money and not supporting those who were clients.

3. Maybe we should require some sort of insurance(and support agreements) for companies who are making medical equipment, esp. implants, so that if somethings like this does happen there is a safety net for the vulnerable people who either volunteered or paid for these things.

4. The words "Move Fast and break things" and xAAS went through my head and made me horrified.

"Hi, we're here to repo your artifical heart. "

There's a movie on that exact topic out there, though u don't recall the name.

Repo Man, with Jude Law and Forest Whittaker. Totally worth it.

Not to be confused with Repo The Genetic Opera, worth it as well, for different reasons. This has Paris Hilton in a bit part, and it's still amazing.

This seems like the obvious end result of letting private startups perform medical care with long term maintenance requirements on people.

Car makers are required to maintain parts availability for 10 years. I think it would be reasonable to have at least similar provisions for bionic parts and service. Say requirements for insurance/trust fund as part of FDA approval. Thus it will be priced in from the start.

Yes, absolutely. We need to push for things like this, at a minimum, if not making everything open source when a company goes belly up who dealt with major breakthroughs like this.

Looks like more ammunition for the right-to-repair movement.

Or to insist on open source for any hardware implanted in my body. Of course, this is often easier said than done.

https://www.thearticle.com/what-happens-when-software-become...

What about prosthesis that are worn, like hearing aids?

IMO, yes. If it aids a disability or some kind of personal problem that someone was born with, inflicted upon themselves, or had inflicted upon them, or had naturally occur, then absolutely.

I also believe that we should put the greater good before profits. Every time.

We should not paywall quality of living, especially the senses.

How has no-one here made a Deus Ex reference yet? It's like a story straight out of the Human Revolution game!

There is a bunch of comments stating that the article is misleading. For example, referring to the first paragraph with the woman describing how her device stopped working. The company's abandoning the tech and almost going bankrupt is only a part of the story, albeit central. They did not disable her device remotely, and they still worked on the tech at the time, but they COULD NOT repair the device:

Barbara Campbell, who received her implant during the clinical trial of the Argus II... But in 2013, after four years of regular use, Campbell’s system shut down in the subway station, and despite some repair attempts by Second Sight, never worked again. While she talked with her doctors about having the implant removed, she ultimately decided that the risks of another surgery weren’t worth it. She still has the defunct technology in her left eye.

It shut down unexpectedly, the company still worked on the tech, yet could not repair it. She now has a defunct neurological device in her head that is supposedly unrepairable and the risk of removing it might outweigh the risk of keeping it. Isn't it terrifying as is? Now to add to the insult, the company that made it shifted to something else, so if it starts causing any troubles, there will be no-one to provide expertise to help assess the risk. It reinforces the main story quite a bit: there is enough risk and complexity as is, regardless the company going out of business.

I thought this was about that app that lets you help blind people by using their camera and you tell them what you see. This is way more sinister. I couldnt imagine depending on something like this and then it's just gone. It would be like having contacts with the worst vision in the world then being like sorry we don't make that kind any more.

But if you say you wish there was less emphasis on splashy headline grabbing medicine like organ transplants and more on helping people keep their "original equipment from the manufacturer" you are a rabble rouser or something.

The Hubris of mortals. Something tells me we are going to see some real horrifying stuff when Silicon Hills moves on to its MVP showcase for resurrecting the dead.

When they do what?

> With all the overhead of sales and regulatory people, it wasn’t profitable

LOL. All the regulation and they didn't protect the patients and they didn't help the company.

If there were no regulations on medical devices, a lot more people would be getting hurt by fraudsters selling snake oil.

It's like grousing about a city's water supply getting contaminated once, and demanding that the remedy is to shut down the water treatment plant. The plant didn't even protect people! We should just drink raw, untreated lakewater!

Unless it turns out that the water treatment plant isn't in fact doing anything but is still a cost, in which case yes, you might as well drink the raw lake water. Perhaps all the people in this town would be better off if they would each boil their own water. It could be that we all end up better off without the treatment plant...

Reply on news.ycombinator.com