I spent years negotiating security audits with global giants. The secret? You don't need to match their spend to match their security.
https://rolandturner.com/Thrift%20Shop%20Security:%20Why%20You%20Don't%20Need%20a%20Bank's%20Budget%20to%20Defend%20Your%20Network/ @fossasia
#InfoSec #BlueTeam #OpenSource #CyberSecurity #NGOTech #ThriftShopSecurity #RiskManagement
I'm willing to assert that *I* am, certainly. I wouldn't dare speak for anybody else...
I'm guessing from your wording that you're not concerned about something specific to something that I've posted (i.e. that would not apply to most other images generated by the same AI), but rather a more general concern about generative AI? If so, then we may have to respectfully differ.
Whatever minuscule reduction that my desisting might contribute to reducing the long list of harms that generative AI is already doing — or is alleged to be doing, or is predicted to do — it would be immaterial. On the other hand, the **only** people who are going to make material contributions to avoiding, preventing, or mitigating those harms are those who have a working familiarity with the technology, which is something that I am consciously pursuing.
(This is not to justify negligent use of course. e.g.:
- Jamming F/OSS projects with AI slop PRs is plainly harmful.
- OpenClaw seems likely a spectacularly bad idea.
- An OpenClaw instance telling the matplotlib maintainers off for declining to accept its PR is probably just the first of a long list of intersecting bad ideas that we'll see over the next few years.
- Etc.
)
Roland Turner
Furbland's Very Cool Mastodon™