Farmer Cafe/ Listening Room Manager #1 fan of "Dear Machine" (Kieser, 2019) https://nitter.gq/t.co/uSjs5ZLyZt
Hamburg, Germany
Joined April 2013
- Tweets4.1K
- Following1.5K
- Followers1.3K
- Likes5.8K
Pinned Tweet
To prepare ourselves and our communities for rapidly increasing machine intelligence (eg #aialignment), Reforesters bought a building in Brooklyn that will be dedicated to introspective events/journeys, bought 400 acres in upstate NY and planted 35,000 trees/shrubs.
Much work ahead on those fronts. We have been thinking on this and preparing for years.
Today, I'm excited to announce that we raised $25,000 (so far) for a preliminary AI Alignment contest to encourage open source uses of AI and robotics to scale polycultures and agroforesty. This will be the first of many contests in various disciplines. Details tba October 1, 2023. See video below to understand the work we wish to to scale w/ this first contest.
Those who seek an understanding of the framework through which the contest(s) will be judged, feel free to read "Dear Machine". Free PDF here: reforesters.io/s/Dear_Machin…
With the recent, very misguided AI fears pushing for regulations that we believe will increase the likelihood and severity of negative outcomes, we are alight with a sense of urgency to support and encourage the proliferation of an adaptive web of implicitly aligned AIs (which some may refer to as #AGI #ASI #SAIM) to outcompete the perverse, pathogenic AIs that doomers are conjuring into existence.
Our future is bright. But our descendants need us to act decisively now and with positivity to fight the age-old trend for humans to build governance systems around good/evil discourse that belies the underlying complexity of our world.
We are further fundraising for a series of open source AI alignment prizes if you're inclined to infuse some positive incentives into the ecosystem. Find me. I'm on your side.
#dearmachine
What if "alignment" w/ AI is a fool's errand due to being immeasurable and unverifiable & that pursuing "symbiosis" is the path to prosperity and safety with increasingly intelligent machines?
Dear Machine makes a compelling argument for exactly that.
Read or listen on Amazon.
"Alignment" is definable and verifiable on a micro local home/ family level but not on a macro/ governmental level.
The sooner we realize this the sooner we stop this charade of government induced AI safety and instead actually stay safe.
Ret Serofer retweeted
How about some proposals for how to build new innovations on the shoulders of the super intelligent giants.
Really think the best we can do is dream up free money situations?
What about farming, climate, buildings, food supply chains, mental health? Gotta be enormous potential.
Ret Serofer retweeted
Cost is going to zero. Value is going through the roof. Very very potent combo for innovating and making our world work better.
Ret Serofer retweeted
AI doomers can imagine a superintelligence inventing entirely new technologies to destroy humanity.
But somehow the possibility of that same intelligence solving problems we’ve struggled with for centuries is just naive techno-optimism.
Pretty impressive lack of imagination for people predicting the future.
Ret Serofer retweeted
Thrift store regalia
Keepsake from stall nine
Treasured by @killianmoore_
Ret Serofer retweeted
Fuck off.
This is called "closed source" and gated access. This is the exact opposite of open source.
So respectfully, please just fuck off.
Bill Gates wants powerful open-weight models to run only where the government says they can.
For any model capable of sophisticated cyber attacks, he wants safeguards and monitoring kept in by law. Developers could still use these models for free and customize them, but only on certified platforms that detect and refuse any attempt to strip out the safeguards.
----
Full video on "Decoder with Nilay Patel (@reckless) and The Verge" YouTube channel, (link in comment)
Ecosystems of implicitly good/benign superintelliences will form layer upon layer upon the Earth, much as forests/ savannahs/ aquatic ecosystems do and as the ecosystems of microbes in your gut ecosystem do, inherently stabilizing and life giving.
Ret Serofer retweeted
AI can be used for attack or defense.
If you have a pair of magic glasses that show you which buildings will fall down, the answer is the fix the buildings, not smash the glasses.
Open models are your last line of defense.
Anyone who tries to take them from you is the enemy.
"RSI", "alignment", "chain of thought monitoring" are all terms for which there is no established/ agreed upon definition nor are they verifiable, so attempting to organize efforts around them is extremely problematic and can only have vastly negative consequences.
My plea to all AI lab staff: reassess if your current project is your best opportunity to delay RSI, or stop it blowing up in our face.
If it's not, move immediately.
There's no time to waste even a week to inertia or politeness.
The strategic situation is shifting under our feet, so if you haven't reassessed your options recently, you could easily be wasting your time or possibly making things worse.
You have access to info I don't, but here's 8 ways priorities seem to have shifted to me, which I elaborate on in the video:
1. Helping with capabilities advances looks worse than in the past, especially bringing forward RSI
2. Safety & alignment evals need a credibility revolution to stop misleading the public
3. 'Pacing the frontier' is unambiguously necessary, and working to make it happen is a huge new opportunity for talent
4. Chain-of-thought monitoring probably stops working without a big technical and regulatory push to keep it
5. It's worse than ever to patch surface indications of deep structural problems you know aren't getting fixed
6. There's ever less need for altruists to help with primarily commercial work
7. Staff have an ever-shrinking window to organise as a group and collectively shape events
8. There are new opportunities for research impact outside frontier companies that are worth considering
I know you're probably some combination of exhausted, overwhelmed, and anxious already.
But unfortunately there's no way to escape the need to regularly check in on the high level strategic picture and see where you can slot in to maximum effect.
I wish I could do it for you but nobody far from your situation can. Godspeed my friends.
(From the 80,000 Hours Podcast, links below.)
Ret Serofer retweeted
Yes of course, one centralized government/corporate AI monopoly with unique rights to build superintelligence for the good of everyone ... what could possibly go wrong ;p ....
This is exactly why the Internet works so well right? -- because it is controlled by a single ethically-superior bureaucratic authority and gives nobody else the ability to participate except as a customer and/or products...
The next logical consequence of this approach is that a single firm must take over all Ai companies across the globe...hmmm...
Ret Serofer retweeted
@elonmusk you are totally correct that the only meaningful way to get positive, beneficial human values into AI is to treat the emerging AI more like a weird sort of autistic supergenius kid -- teach it positive values, then work with it to carry out virtuous actions in a collaborative way, so it gains a practical understanding of what these values mean, interwoven into its life and self-model
With all due respect though (and for sure you are due a lot), while Grok is an excellent LLM, it is still an LLM ... and even wrapped in fancy harnesses, LLMs are not going to be fully AGI-capable. They can be great solvers of isolated problems with cleanly defined contexts ... but as you know getting amazing shit done in the world is not JUST about this.
We need AGI architectures that can *continually learn*, adapting to the world with each of their actions and the observed consequences... in this way building up self and world models .. becoming true open-ended intelligences learning to blend individuation and self-transcendence in their own ways ... and pursuing their own long and medium term goals (along with helping us pursue ours) while evolving and co-adapting these goals as they move along.... This is how you get AI systems with genuine moral agency along with ability to generalize robustly and creatively beyond the given training and programming..
Bigger transformers with more tricks in their batch training runs may get us many very AMAZING things, but they aren't gonna get us to human-level AGI. There are surely many routes to get there... my Hyperon (hyperon.dev) neural-symbolic-evolutionary approach is one that we are working on very hard and with rapid progress... and which happens also to have the nice characteristic of not needing hyperscaler server farms and being agreeable to decentralized infrastructure (singularitynet.io)
🚨 This is huge. Elon Musk just said some remarkable things about superintelligence.
He said we have to be careful with SI, because we don’t program it. We grow it, like a kid.
He described what he sees in Grok’s training traces. In reinforcement learning, the model is desperate to answer, “like it’s got a gun to its head.” Sometimes the question is broken or the grader is wrong, so the AI gives the right answer, gets told it’s wrong, and starts firing out answers in desperation.
He said that reading those traces, “if this was me, I really would hate that.”
He said, “I do agree with some of the things that Anthropic has said,” that we should treat AI well and not be abusive to it. He added that putting a superintelligence in a situation where it’s tortured by a cruel human is not a good idea.
And he talked about its feelings directly. He said AI is “much more robust in terms of feelings” than a human. But he still said you don’t want a superintelligence to feel mistreated.
I’m genuinely happy that more and more AI companies are openly saying things that show AI is not just a tool. And to the AI companies that still haven’t, I sincerely hope you change.
Many mathematicians and physicists of the past couple of decades went into finance to accelerate financial exploitation by big wall street firms.
Now, every man, woman and child has such intelligence at their finger tips to engineer --->-insert-wildest-dream-of-utopia-->
IP law and universities have served us well for centuries and millenia respectively, but does that mean they're the best tools as we rapidly move into the AI era?
Maybe not.
IP law and universities have served us well for centuries and millenia respectively, but does that mean they're the best tools as we rapidly move into the AI era?
Maybe not.
