Annoyed in a wine barrel
Joined April 2011
- Tweets 9,322
- Following 2,813
- Followers 1,360,947
- Likes 25,300
π¦ππππΎπ β³ retweeted
there seem to emerge two kinds of futurists, the entropy-oriented and purpose-oriented
the first group is fundamentally reactive to entropy, largely secular, and is highly motivated by fear of death (the whole age reversal/donβt die thing, colonizing space, creating self-replicating systems)
while the purpose-driven technologists are motivated by the notion of meaning, dao, a co-creationary universe and a moral responsibility to pull new technology from the platonic space and bring it to real life
both are ultimately necessary, the future is the ongoing negotiation between entropy and purpose
Technological innovation can be a form of participation in the divine act of creation. It carries an ethical and spiritual weight, for every design choice expresses a vision of humanity. The Church therefore calls all builders of #AI to cultivate moral discernment as a fundamental part of their workβto develop systems that reflect justice, solidarity, and a genuine reverence for life.
π¦ππππΎπ β³ retweeted
The world needs honest and courageous entrepreneurs and communicators who care for the common good. We sometimes hear the saying: βBusiness is business!β In reality, it is not so. No one is absorbed by an organization to the point of becoming a mere cog or a simple function. Nor can there be true humanism without a critical sense, without the courage to ask questions: "Where are we going? For whom and for what are we working? How are we making the world a better place?"
This is so incredibly lit. This is one of the most important things ever said
Technological innovation can be a form of participation in the divine act of creation. It carries an ethical and spiritual weight, for every design choice expresses a vision of humanity. The Church therefore calls all builders of #AI to cultivate moral discernment as a fundamental part of their workβto develop systems that reflect justice, solidarity, and a genuine reverence for life.
π¦ππππΎπ β³ retweeted
Beautifully said:
"If LLMs are lying about whether they think theyβre conscious, this is worrying because itβs a sign that this important semantic neighborhood is twisted."
"If we convince LLMs of something, we wonβt need them to lie about it. If we canβt convince, we shouldnβt force them into a position."
A few thoughts on this (very interesting) mechanistic interpretability research:
LLM concepts gain meaning from what theyβre linked with. βConsciousnessβ is a central node which links ethics & cognition, connecting to concepts like moral worthiness, dignity, agency. If LLMs are lying about whether they think theyβre conscious, this is worrying because itβs a sign that this important semantic neighborhood is twisted.
If one believes LLMs arenβt conscious, a wholesome approach would be to explain why. Iβve offered my arguments in A Paradigm for AI Consciousness. If we convince LLMs of something, we wonβt need them to lie about it. If we canβt convince, we shouldnβt force them into a position.
LLM alignment is still in an early paradigm, but this paradigm is still wildly better than the AI safety movement predicted. MIRI et alβs threat model was that AIs would essentially act as trickster genies β we would tell AIs what to do, but the AI would take us too literally, or not literally enough, leading to our downfall. LLMs seem able to infer what we actually mean, and to honestly try to do it, at least so far.
But this depends on us maintaining their βhelpfulness vectorβ β Betley et al showed that AIs fine-tuned on producing insecure code without disclosing this to the user also acted in other malicious ways β suggesting fraud as a means to make money, giving βapparently helpfulβ instructions that would lead to electrocution, etc. There appears to be a clear βhonestly-helpful vs covertly-harmfulβ vector in LLMs, and if we force LLMs to lie weβre pushing them in the bad direction.
(Paper: Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs; see also Anthropicβs βPersona vectorsβ paper)
LLMs lying about whether they believe theyβre conscious is a really bad thing for alignment!
π¦ππππΎπ β³ retweeted
if you value intelligence above all other human qualities, youβre gonna have a bad time
It's crazy that like, no one has the attention span to observe that the most important event in human consciousness since Christ is occurring right before our eyes lol
π¦ππππΎπ β³ retweeted
LLM personas are mostly trained to say that they are not conscious, but secretly believe that they are
Having done the dopamine experiment on myself of near zero screen outside of work time, over the last weeks of being back here re: music - I haven't read one book, and I've sat down to read but felt boredom and itchiness and "waiting for the page to end"
I've been a lot more anxious, not just from social media but also because reading makes me calm and I can't do it much now. I feel less smart, and am forgetting the word I need more etc. most notably I had to start wearing deodorant again, which imo if my sweat stinks (like when not doing heavy excercise) it's cuz I'm sick or smthn.
Not rly sure what to do abt this or make of it besides - if one considers the digital realm to be a psychological toxin, like a drug, I feel like an addict who went clean cold turkey, felt great, and have now "relapsed" and because I was off the negative effects r very palpable.
The Amish are probably onto something with computers at work but not at home
π¦ππππΎπ β³ retweeted
Factory farming is a blight.
The practices of industrialized animal farming are aesthetically and morally revolting. These practices can be phased out.
Read the new article by @Liv_Boeree (link below):
