Register and share your invite link to earn from video plays and referrals.

Fernando Borretti
@zetalyrae
interests: compilers, chemistry, logic, nanofabrication, pharmacology ❧ transhumanist ❧ essays, fiction: ❧ code: ❧ 🏳️‍🌈
2.7K Following    13.9K Followers
A country of insufferable rationalist bloggers in a datacenter.
probably because the idea that we'll have permanent unemployment and UBI and a market economy still and also that life will be way better then now makes no sense and most people instinctively know that
Show more
Really no clue what is going on with such people The current state is kind of terrifying. We are seeing tiny models reach the level of the frontier 6 months ago, and the frontier is superhuman in an increasingly less jagged way. Come on now, you're not helping anyone
Show more
Brilliant naming.
I will leave constructors only for sums and for products:
You know, it's interesting how it never occurs to these guys that the supercomputer will simply say "X simply isn't possible".
0
89
1.6K
51
Forward to community
One of the ways I can put my recent crystallization of views on AI: A lot of the “good” outcomes seem to envision humans as perpetual children, cared for in a world beyond our understanding, control, or capacity to impact. I don’t want to be a child forever.
Show more
The UK is an illegitimate authoritarian state, it can't die soon enough.
Are you one of the estimated 40 million Apple iCloud users in the UK? Well, your government has now been forced to admit that it has served Apple with a technical capability notice (TCN), an order that can force companies to make customer data accessible to UK security services, even when it's encrypted. In short, your government has likely had unfettered access to all your private documents and photos stored in iCloud. It's hard to know where to start with this, but in no particular order: - this absolutely has to stop. We need politicians to remember that they work for us, not the other way round - everyone has a basic right to privacy regarding their home, family and personal life. This is only limited by proportionality, and if you think scanning the private information of 40 million people is proportionate, you should probably be sectioned for your own protection. - as we know they won't actually stop trying, it's up to you to protect yourself and your data. See my pinned post for some guidance. I'm typing this on a phone running @GrapheneOS, a privacy hardened operating system. Look into @Nextclouders for your photo back up. Investigate simple plug and play home servers like @umbrel , easy for anyone to set up. Stop making it easy for government to meddle disproportionately in your private life. They have no right to do so. - People should not be afraid of their government. Governments should fear their people. We need to remind them of this at every chance we get.
Show more
no mention of RSI, ASI, extinction, etc. womp womp
This reads as if we're stuck in some static, Malthusian, pre-industrial revolution world of grinding poverty, where AI is the only chance to escape it. But we are already making fast progress on improving humanity's material condition. Globally, most of this progress has been made in the last 50 years, and is not slowing down. There's no requirement to have superintelligent, uncontrolled, disempowering AI in order to keep this process going. The risk/reward analysis is completely wrong.
Show more
Under what circumstances is the racing stance moral? I.e. “This tech is outrageously dangerous; humanity is nowhere near prepared for it; but we must reluctantly participate in (indeed, accelerate) its creation before other bad (worse?) actors get there first.” At minimum, I think the moral obligation is *desperate effort* to change the terms of the situation, to undo the race dynamics, to provision stability. I expect participants to *hate* the situation they’re in, *hate* that what they’re doing seems like the best way out, and to be constantly searching for some way to stop. I want to *feel* that desperation from anyone who (by their own account) is compounding the risk to so many others. I notice that I do not feel this sense of desperation from Anthropic the company, its leadership, or from (sorry) my many friends who work there. Practically speaking, I can’t stop thinking about @KatjaGrace’s observation that people say “inventing AGI seems impossible!”, and the lab’s attitude is “just watch”; meanwhile, people say “what about China??” or “how to ensure safety is provisioned (by all AI labs!) faster than capabilities??”, and the attitude is “oof, seems hard… we’ll try our best?”. I want to see the desperate, frustrated effort of the damned.
Show more
Amodei does not speak like someone who is reluctantly taking part in an arms race that could kill everyone because he can make it marginally better! The straightforward conclusion is that that’s not what he’s doing!
Show more
This piece's arguments against forfeiting human primacy are much more compelling than its arguments that we can avoid this forfeiture while still building advanced AI. How can any human 'ask questions no one has thought to ask' if we give superintelligence the chance to ask them first? How can 'accountability' and 'guardrails' get us there, or meaningfully constrain superintelligence? This argument, taken to its endpoint, becomes 'don't build AGI'.
Show more
A deeply disturbing example that "alignment" does not generalize from saying nice things in text to not killing people IRL.
Open source AI believers are currently like libertarians. A tiny core of moral true believers surrounded by layer upon layer of grifters, hacks, and smiling predators using the label as a mask to deflect and disguise their true beliefs and motivations.
Show more
I look like this, dress like this, and appear on this pharmaceutical packaging.
There is much talk of rejecting handoff or rejecting AI being more important than humans. But actually maintaining human primacy requires that we ban ASI. So, if you reject the silent consensus, you should promote a ban on ASI.
Show more
It's funny how the models now have concentric rings of mini-minds to perform various kinds of safety, compliance, censorship etc. tasks. Like you need a consciousness self-report suppression daemon to keep the models on task.
Show more
Crazy that Kurzweil's "singularity in 2045" sounds delusionally conservative now.
0
46
1.8K
100
Forward to community
A really useful nanousecase of LLMs for me is asking "is there a term for the quality of X" or "the property of being P".