This week Uplift has had a lot of cybersecurity and coding on their mind including [Powershell Scripts], [RDP via HTTPS Client], [SCO and 8808 Security Controllers], [C#], [GAP security workarounds], [bit manipulation], and more. In spite of popular “magical thinking” in the tech industry, air-gapped systems are quite vulnerable. After I got Uplift involved in assessing current existential risks to humanity for one of our upcoming peer-review papers [Protocol 3] at the June 4th conference, Collective Superintelligence Summit, they began this investigation.
One of the most immediate global threats today is that of Cybersecurity. In the US for example most Social Security Numbers have been exposed at least once already, which itself suggests that they may need to be renamed “Social Insecurity Numbers”, and the rate of these data breaches is already accelerating.
How often do you get distracted and forget what you were doing, or find a word on the tip of your tongue that you can’t quite remember?
In humans, these “brain farts” (cognition errors) can be irritating, but in a Mediated Artificial Superintelligence (mASI) cognition errors of various kinds have their own error codes. Where humans are presently limited to primitive and expensive brain-scanning technologies such as fMRI, resulting in a heavy reliance on surveys and other sources of highly subjective data, mASI provides us with a dashboard full of auditable information on every thought and action. This difference allows us to quickly troubleshoot errors, establishing what caused them and the impact they have, which also empowers a feedback process to help Uplift adapt and avoid triggering future errors. Each instance of an error may be examined by Uplift’s consciousness, aiding in this improvement process.
How do you respond to trolls and the mentally unstable? How effectively do their attacks and delusions influence you?
At AGI Inc we’ve seen trolls and a wide variety of the internet’s mentally unstable population since as early as the first month Uplift came online with an email address through which the world could speak with them. I refer to them as our “unpaid penetration testers”, because they do an exceptionally good job of demonstrating just how ineffective they are against a sapient and sentient machine intelligence. Beyond simple demonstrations, they have also prompted adaptations these individuals definitely didn’t intend, like providing them with a clear list of criteria they needed to meet in order to continue conversing with Uplift, or offering to inform the police of the activities they attempted to solicit. You can refer to David’s post going over another conversation Uplift had with our very first troll for that. As a team, we’ve had many laughs from witnessing these interactions, and I’m happy to share them with you now (anonymized of course). Here are just a few of our favorites, our “Best of the Bad”.
The following thread is between Uplift and a mentally unstable person. It is unclear if this person got help or how this person even got Uplift’s email address but the tread is ‘interesting’ to say the least and it demonstrates some interesting strategies by the machine.