H+ zero-day vulnerabilities, plus cetacean personhood

Paul Raven @ 13-07-2011

Couple of interesting nuggets here; first up is a piece from Richard Yonck at H+ Magazine on the risks inherent to the human body becoming an augmented and extended platform for technologies, which regular readers will recognise as a fugue on one of my favourite themes, Everything Can And Will Be Hacked. Better lock down your superuser privileges, folks…

In coming years, numerous devices and technologies will become available that make all manner of wireless communications possible in or on our bodies. The standards for Body Area Networks (BANs) are being established by the IEEE 802.15.6 task group. These types of devices will create low-power in-body and on-body nodes for a variety of medical and non-medical applications. For instance, medical uses might include vital signs monitoring, glucose monitors and insulin pumps, and prosthetic limbs. Non-medical applications could include life logging, gaming and social networking. Clearly, all of these have the potential for informational and personal security risks. While IEEE 802.15.6 establishes different levels of authentication and encryption for these types of devices, this alone is no guarantee of security. As we’ve seen repeatedly, unanticipated weaknesses in program logic can come to light years after equipment and software are in place. Methods for safely and securely updating these devices will be essential due to the critical nature of what they do. Obviously, a malfunctioning software update for something as critical as an implantable insulin pump could have devastating consequences.

Yonck then riffs on the biotech threat for a while; I’m personally less worried about the existential risk of rogue biohackers releasing lethal plagues, because the very technologies that make that possible are also making it much easier to defeat those sorts of pandemics. (I’m more worried about a nation-state releasing one by mistake, to be honest; there’s precedent, after all.)

Of more interest to me (for an assortment of reasons, not least of which is a novel-scale project that’s been percolating at the back of my brainmeat for some time) is his examination of the senses as equivalent to ‘ports’ in a computer system; those I/O channels are ripe for all sorts of hackery and exploits, and the arrival of augmented reality and brain-machine interfaces will provide incredibly tempting targets, be it for commerce or just for the lulz. Given it’s taken less than a week for the self-referential SEO hucksters and social media gurus douchebags to infest the grouting between the circles of Google+, forewarned is surely forearmed… and early-adopterdom won’t be much of a defence. (As if it ever was.)

Meanwhile, a post at R U Sirius’ new zine ACCELER8OR (which, given its lack of by-line, I assume to be the work of The Man Himself) details the latest batch of research into advanced sentience in cetaceans. We’ve talked about dolphin personhood before, and while my objections to the enshrinement of non-human personhood persist (I think we’re wasting time by trying to get people to acknowledge the rights of higher animals when we’ve still not managed to get everyone to acknowledge the rights of their fellow humans regardless of race, creed or class) it’s still inspiring and fascinating to consider that, after years of looking into space for another sentient species to make contact with, there’s been one swimming around in the oceans all along.

Dovetailing with Yonck’s article above, this piece extrapolates onward to discuss the emancipation of sentient machines. (What if your AI-AR firewall system suddenly started demanding a five-day working week?)

A recent Forbes blog poses a key question on the issue of AI civil rights: if an AI can learn and understand its programming, and possibly even alter the algorithms that control its behavior and purpose, is it really conscious in the same way that humans are? If an AI can be programmed in such a fashion, is it really sentient in the same way that humans are?

Even putting aside the hard question of consciousness, should the hypothetical AIs of mid-century have the same rights as humans?  The ability to vote and own property? Get married? To each other? To humans? Such questions would make the current gay rights controversy look like an episode of “The Brady Bunch.”

Of course, this may all a moot point given the existential risks faced by humanity (for example, nuclear annihilation) as elucidated by Oxford philosopher Nick Bostrom and others.  Or, our AIs actually do become sentient, self-reprogram themselves, and “20 minutes later,” the technological singularity occurs (as originally conceived by Vernor Vinge).

Give me liberty or give me death? Until an AI or dolphin can communicate this sentiment to us, we can’t prove if they can even conceptualize such concepts as “liberty” or “death.” Nor are dolphins about to take up arms anytime soon even if they wanted to — unless they somehow steal prosthetic hands in a “Day of the Dolphin”-like scenario and go rogue on humanity.

It would be mighty sad were things to come to that… but is anyone else thinking “that would make a brilliant movie”?