Artificial Intelligence Risks Open Source and the Internet

Artificial intelligence
Facebooktwitterredditlinkedin

            Houston           If we don’t have enough to worry about, there’s always more, right?

Organizationally and personally, we’re fans of open source as our choice for access to email and internet.  I should add to databases for ACORN members as well, since we use Civi.  Why support the tech overlords anymore than we must?  That’s been our operating philosophy.

What am I talking about now, fellow techno-peasants?  Proprietary systems are those controlled by Microsoft, Facebook, Google, and the rest.  They might offer you access for free, but they are making bank off your information and personal coordinates.  They control the code that powers the operating systems.  Open source on the other hand makes the code transparent and available, so anyone can modify the system to suit their purposes.  When you do so, you also advance the whole community, because they can also see the changes you have made, and, if they like them, adapt them to their operations as well.  Open source is also free.  It’s not a perfect system, but it’s more than good enough and the politics are immensely better.

In this spirit, I support Mozilla, which is an open-source internet access system.  They also provide me with my email server system, Thunderbird.  Organizationally, we have our own server system, but when it all comes home to me, it’s all thanks to Mozilla.  They are also nonprofit, which is something else worth supporting.  I dutifully make a small contribution at the end of the year when they ask for support, because I depend on them and they need help from all of us.

All of which made me pay close attention to an op-ed in the Times written by Mozilla’s chief technology officer with the catchy title:  “It’s the End of the Internet as We Know It” – what the frick!?!  This is about the unrestrained power of artificial intelligence.  If you missed it, one of the leading AI operators, that many think is better than most of the bunch, Anthropic did a limited release of its new model, Claude Mythos Preview.  Allegedly, it’s so powerful they ended up only putting it out to 50 of the world’s largest outfits, because it is able to find security glitches in software.  Others are also developing something similar.

At first this sounds good.  Hey, they’ll be able to increase cybersecurity.  Yes, among the 50 that are part of Project Glasswing, but not for the rest of us.  Here’s the rub as well.  Many are getting AI to write code for special purposes in their outfits. Others, like ourselves, may use or intersect with those programs and systems.  These on-the-cheap operations don’t have the same cyber controls or top engineers cleaning their code obviously, so hackers will be having a heyday, ending the internet as we know it, then all our stuff is gone pecan.

Mozilla makes a couple of indisputable points.  One is that open-source software built the internet and is still the infrastructure under streaming, firewalls, and other services built over the years by nonprofits and volunteers.  The other point they are making is that they think these AI firms need to “contribute engineering time, security expertise, and staff to the projects we all depend on.”  In other words, they need to do no harm, and protect all of us, not just their insider club of big whoops.

Who can disagree?  If you are the folks opening Pandora’s box, then you sure as heck need to help close it!

 

Facebooktwitterredditlinkedin