AMA with Jeffrey Ladish. Wednesday 2/4 at 2:00 PM CT
Executive director of Palisade Research; studying AI loss of control risks.
AMA with Rob Miles on AI Safety. Wednesday, 2/4 at 1:00pm CT
AISafety.info founder has spent years telling the world about risk posed by strong AI.
https://www.youtube.com/watch?v=6tYqqb6AjTMAMA with Nate Soares. Wednesday 2/4 at 10am CT
Author of If Anyone Builds It, Everyone Dies answers questions about why superhuman AI would kill us all.
AMA with Nate Soares. Wednesday 2/4 at 10am CT
Author of If Anyone Builds It, Everyone Dies answers questions about why superhuman AI would kill us all.
What is the 'Metacrisis' and How Do We Solve It? (AMA). Rewatch the live AMA conversation with Layman Pascal
https://www.youtube.com/watch?v=jyq_ZfdtTmgMy Journey with Claude Code. The more I use Claude Code, the more impressed I become.
I keep throwing progressively harder problems at it, and whenever the problem is conceptually tractable, it can usually just solve it. Not with hacks or brittle workarounds, but by actually engaging with the structure of the problem.
I decided to push it further by combining two difficult problems.
Difficult in the sense that either one would likely take me years to complete properly on my own let alone together whether the interaction between the two is a kind of complexity in its own right.
With the exception of one genuinely catastrophic error that required intervention to recover from, the tool has kept going, iterating, and making real progress.
What stands out most is that it seems to understand what progress actually is.
It does not treat the number of passing tests as a sacred metric. It is willing to break tests if that moves the system forward in a deeper, more honest way. That is something many humans struggle with.
The mere fact that it can reason about progress at all, rather than optimising a superficial proxy for it, is pretty remarkable
And to think the tool over past few months has seen pretty consistent improvement at the cadence of weeks with no end in sight.
Will software development be unrecognisable a year or even 6 months from now, I do not know.
Isn't It Ironic? . Don't you think?
A little toooo ironic.
I really do think.
This site where we are upping trust lets users post under a pseudonym.
Every time I see a post or comment from a pseudonym, screen name, handle, what have you, after first wondering if it's another godforsaken AI bot stealing my eyeballs away from human creations, I remember a line from the Tao Te Ching;
To give no trust
is to get no trust.
v.17, Lao Tzu x Ursula LeGuin
If I'm stunned by Claude's clarity and insight in response to my posting Nithya and my takes on Atisha... Is that a reflection of Claude? Me? Nithya? Atisha? Humanity’s wisdom? Something else? (OP)
https://uptrusting.com/post/dB735BExperiment: How is whatever's happening serving the greater good? If we zoom out long enough, we can often see that massive setbacks created foundations for evolution. Eg:
What’s a thing in the world that you don’t like right now, and think is a huge step backward, that might also be a step forward? How so?
By design, this is an unverifiable experiment from a third person perspective. Since we can keep zooming out + everything is interconnected, we’ll probably never know for sure, even if we live for thousands of years.
But by design, this is verifiable from a first person perspective: Does your experience improve or change in any way by the experiment? How so?
(note that this doesn't ask you to deny any suffering—such as the horror of the oxygenation event's great extinction, or stop trying to make things better. Like everything, this perspective can be misused. "Everything happens for a reason" is usually dismissive, "if there were a reason for this in the long run, what might it be?" is additive. Like allowing versus expressing, it's not about bypassing the difficulty but rather creating a larger container for it. Freedom comes through acceptance rather than resistance.)
#TTT
also, um, what are we supposed to make of the AI’s image? That’s the representative of human error
?
Web Discussions: Flat by Design (2012). web archive (original URL is broken)
A web discussion about threaded versus linear web discussions. I recommend reading the article and many of its comments. I think the two approaches aren’t mutually exclusive and that the best design will be something that includes both. Our tree view
is an experiment in combining the two.
Who should date me? I get really pumped about what can be possible for online dating in UpTrust.
For example, someone having high current trust scores from a majority of his exes.
Or me going through all my single friends who are the gender I date and posting on their Dating Recommendations
tree posts like I adore him as a friend and would totally want to date him if we wanted the same lifestyle! He gives the best hugs and is someone whose opinion I value highly when I make major decisions. I’d like to see him with a woman who embraces spirituality and likes throwing Superbowl parties.
So here’s my personal test-drive. Feel free to populate this tree with your opinions on my dating life, recommendations, questions, etc. <3