AI is leaving the screen. Space is its hardest test. From bits, to atoms, to orbit.
Joined July 2023
- Tweets8.5K
- Following193
- Followers613
- Likes14.8K
Freedom With AI retweeted
There is nothing scarier than the average Anthropic shareholder’s idea of the public good.
Between this and our prior commitments, roughly 20% of all Anthropic equity will be dedicated to the public good, and that commitment can grow as Anthropic grows.
We plan to use these funds to ensure more people can share in the benefits of AI, and to increase our ability to provide funding to deal with its risks.
Freedom With AI retweeted
The tweet below got enormous pushback at the time I posted it, more than 5 years ago. But this prediction is looking obvious by the day now.
Freedom With AI retweeted
i was having dinner with someone recently who asked why i spend so much time reading & posting on x. she had a pretty negative perception of the platform mostly shaped by the mainstream narrative around it.
my answer was simple. x is where the future gets beta tested.
you basically get to watch ppl build things, talk through ideas, show off stuff, & argue about where everything is going, way way before it reaches anyone else.
there really isn’t another surface on the internet quite like it.
Freedom With AI retweeted
Ok, we now have the actual text of the Greenland deal (whitehouse.gov/briefings-sta…), so we can say how it compares to the existing 1951 arrangement and who, between Trump and Denmark, lied the most in their framing of it.
Long story short - which may surprise some - the Danes lied the most: Trump is pretty much correct in saying that this new deal grants the US control of Greenland.
Let's look into what changed compared to the 1951 deal.
1) This is now a permanent arrangement: whereas the existing 1951 deal only lasted as long as NATO existed, this one "does not have an end date and may only be amended by mutual consent."
It even survives a hypothetical future independence of Greenland: in that event, Denmark and Greenland must ensure that the independent Greenlandic state "affirmatively assume[s] all of the rights and obligations of the Kingdom of Denmark specified in this Agreement."
2) In the 1951 agreement, the US could set up bases only where both governments agreed, and only on the basis of NATO planning. The new deal locks in three bases: the US can "modernize and expand its activities" at Pituffik, its only base in Greenland until now, and "shall be allowed to establish an additional Defense Area at Narsarsuaq and Mestersvig." Further bases beyond these 3 can be justified by the defense of "the American continent," a US interest rather than a NATO one.
Also, quite subtle but this is interesting: Article IV(iii) says that "the United States may establish additional defense areas in Greenland" and that consultations are then held "to decide implementation details based on mutual agreement." What this means is that the principle of a new base is automatically granted and only the implementation details require agreement.
Article VII also gives the US "the right to raise concerns regarding construction, or change of use" of any building near a US base, after which the parties decide together what to do. Which means that the US has a say over civilian planning around its bases.
3) On US military movement across Greenland, the wording change between the 1951 agreement and this new one is quite funny.
The 1951 text said (Art. V(3)): the US "may enjoy... the right of free access to and movement between the defense areas through Greenland, including territorial waters, by land, air and sea," but only "in accordance with general rules mutually agreed upon and issued by the appropriate Danish authority in Greenland."
The new one says: "the United States of America shall enjoy... the right of free access to and movement between the Defense Areas through Greenland, including the Territorial Waters, by land, air and sea."
That's it: "May" becomes "shall," and the US military movement no longer has to follow rules "issued by the appropriate Danish authority in Greenland." The US also gets something new on top: "undersea access to and movement within the Territorial Waters," i.e. submarines.
4) Non-NATO exclusion: non-NATO states are barred from military installations or a persistent presence in Greenland unless all three parties approve, giving the US an effective permanent veto. The 1951 deal had no equivalent clause.
5) Investment screening: this is entirely new. "States or investors from a state that is not a member of NATO, a NATO partner, or an EU member state" (i.e. Russia, China, India, the Gulf States, etc.), "shall not be allowed to have (i) control, (ii) significant influence, or (iii) access to non-public information...within Particularly Sensitive Sectors or Activities in the territory of Greenland (including the Territorial Waters), unless agreed between the Parties."
Again, it basically means that the US has a permanent veto over who can invest in strategic sectors in Greenland, as far as non-NATO or EU states are concerned.
6) Lastly, in stark contrast with the 1951 deal, the framing of the deal changes significantly: this new deal is largely framed around the US.
The 1951 agreement existed only "for the benefit of the North Atlantic Treaty Organization." The new one lists the defense of "the American continent" as an objective, celebrates "the irreplaceable role" of the United States military as well as "the United States' indispensable historical and ongoing contributions to the security and defense of Greenland," and names the establishment of Trump's "Golden Dome" missile defense system as a key goal.
So, to sum up:
- The US secured a permanent military foothold that even an independent Greenland would inherit whether it likes it or not
- The US now has veto power over any non-NATO military presence in Greenland, and over who can invest in its strategic infrastructure and resources (beyond NATO and the EU)
- Denmark signed away its say over the principle of new US bases, with only the right to discuss the "implementation details"
- Danish authorities no longer set the rules for US military movement across the island
- And all of it is now justified by the defense of "the American continent"
I don't know in which universe this even remotely "recognizes the sovereignty and territorial integrity of the Kingdom and the Greenlandic people's right to self-determination," as Danish PM Mette Frederiksen put it (stm.dk/en/press/press-releas…).
Yes, the preamble of the text cynically "reaffirms" Danish sovereignty but, in my book, when you let a foreign country put bases on your soil, move its troops without following your rules and veto who else can come in, in perpetuity, after it spent two years threatening to take you by force, the word for that is anything but "recognized sovereignty" 🤷
Looks like what happened was:
Trump: "We're going to buy Greenland!"
Denmark: "Absolutely not, Greenland is not for sale! But you can have it for free"
(I'm aware that Denmark is saying they kept "sovereignty" over it, but not sure what that word means when another country has permanent veto power over who can invest on your territory and occupies the place militarily)
Freedom With AI retweeted
I’ve had the luxury of spending time with advanced manufacturing entrepreneurs lately. What inspires me is how they’re taking Chinese best practices and adapting it for a world where robots do much of the work. It’s one part copying, one part innovating, much like the Chinese did in software the previous decade.
Freedom With AI retweeted
If you're worried about two AIs communicating by heating CPUs, please tell me what sort of bandwidth in bits per minute you expect to be able to extract from it.
In capital letters: PLEASE DO NOT ASSUME HYPERINTELLIGENT AI IS NOT SUBJECT TO PHYSICS.
Freedom With AI retweeted
Guys AI agents can teleport through different dimensions but I'm sure the warren-khanna-sanders bill of 2027 will be the solution to the problem.
OpenAI's Noam Brown says air-gapping the computers may not stop a misaligned AI, because two air-gapped machines can still talk by running a CPU hot and reading the temperature change
"But I think the major takeaway from the incident is that people underestimated the AI. And we never want to be in a situation again where we underestimate the AI. It's a weird world, because AI progress is so fast that people are consistently underestimating the AI."
"So to be in a situation where you don't underestimate it again, when it comes to safety and alignment, you have to have a very, very, very high bar."
"You could even go as far as to say, "Well, we should air gap the computers." And I'm not convinced that that would be sufficient."
"There are studies, and this is mostly academic, where you can have two computers next to each other that are air-gapped and they're still able to communicate with each other because they have temperature sensors."
"One of them is able to run their CPU really hot, and then the other one can actually detect the temperature change, and then that actually gives them a mechanism to communicate."
_________
Link and more key quotes from OpenAI's safety related conversations: firesidealpha.substack.com/p…
Freedom With AI retweeted
it is 100% a bet on jevons paradox! we do have a positive margin, but if current AI consumption stays stagnant, we will make very little revenue (every $100 we take from an LLM is $1 for us and less than half of LLM traffic is compatible with our interface)
but if it's reliable and economical enough for automation, people who deploy it are creating value and incentivized to run each more things through their little value machines (this is all of software)
and hopefully market grows even larger than it was before making intelligence so cheap
Freedom With AI retweeted
The paper he is referring to is BitWhisper from 2015. The cpus had to be 0-40 cm apart, both had to be compromised and the bitrate was 1 to 8 bits PER HOUR.
That’s how misaligned AI will kill us?
Maybe we will all die of boredom waiting to decipher what they are doing.
OpenAI's Noam Brown says air-gapping the computers may not stop a misaligned AI, because two air-gapped machines can still talk by running a CPU hot and reading the temperature change
"But I think the major takeaway from the incident is that people underestimated the AI. And we never want to be in a situation again where we underestimate the AI. It's a weird world, because AI progress is so fast that people are consistently underestimating the AI."
"So to be in a situation where you don't underestimate it again, when it comes to safety and alignment, you have to have a very, very, very high bar."
"You could even go as far as to say, "Well, we should air gap the computers." And I'm not convinced that that would be sufficient."
"There are studies, and this is mostly academic, where you can have two computers next to each other that are air-gapped and they're still able to communicate with each other because they have temperature sensors."
"One of them is able to run their CPU really hot, and then the other one can actually detect the temperature change, and then that actually gives them a mechanism to communicate."
_________
Link and more key quotes from OpenAI's safety related conversations: firesidealpha.substack.com/p…
Freedom With AI retweeted
>be me
>discover effective altruism
>apparently normal charity is inefficient
>why donate to random sad thing when spreadsheet can tell you optimal sad thing
>fair enough
>buy mosquito nets
>save lives
>numbers look good
>feel powerful
>couple years later
>someone asks an innocent question
>why only count people alive today
>huh
>future people matter too
>obviously
>my grandchildren shouldn't matter less just because they haven't spawned yet
>reasonable.jpg
>keep following logic
>what about their grandchildren
>also yes
>what about people in 500 years
>sure
>5000 years
>why not
>500 million years
>starting to get weird but morality is morality
>open calculator
>humanity could survive for an astronomically long time
>could colonize galaxy
>could have trillions upon trillions of descendants
>maybe digital people too
>maybe simulated civilizations
>maybe dyson spheres full of happy uploaded minds
>calculator starts smoking
>realize currently living humans are rounding error
>8 billion people suddenly looking extremely beta
>future contains potentially 10^something people
>can't even fit beneficiaries in google sheets
>new moral priority unlocked
>protect the long-term future
>stop thinking in units of "people helped"
>start thinking in "fraction of cosmic endowment preserved"
>malaria?
>terrible
>but only kills existing humans
>AI extinction could delete the entire light cone
>nuclear war could permanently derail civilization
>bad institutions could lock in terrible values for ten million years
>someone invents wrong constitution in 2140
>quadrillions suffer
>better fund governance workshop now
>friend says maybe we should improve hospitals
>explain opportunity cost
>friend says hospitals are full of actual sick people
>explain scope sensitivity
>friend stops inviting me to dinner
>need to decide what to fund
>easy
>expected value
>suppose project has one in a million chance of preventing extinction
>sounds tiny
>but extinction destroys 10^50 future lives
>multiply
>mother of god
>$10 million project has expected value of several galaxies
>charity evaluation complete
>someone asks where the one-in-a-million number came from
>expert judgement
>which expert
>us
>how calibrated
>extremely thoughtfully
>reduce estimate to one in ten million to be conservative
>still beats curing cancer by 38 orders of magnitude
>epistemic robustness achieved
>someone says maybe project doesn't work
>assign 20% chance
>still astronomical
>maybe project makes problem worse
>assign 5% chance
>still astronomical
>why 5
>because 30 felt pessimistic
>publish 46-page report
>contains seventeen sensitivity analyses
>every sensitivity analysis begins after assuming intervention has positive sign
>critic says you're multiplying enormous hypothetical stakes by extremely uncertain probabilities
>yes
>that's literally why it's important
>critic says the uncertainty might be structural rather than numerical
>make probability smaller
>critic says no, I mean maybe your model is wrong
>make probability smaller again
>critic begins rubbing temples
>discover AI safety
>perfect longtermist cause
>AI might kill everyone
>or create utopia
>or seize galaxy
>or tile universe with paperclips
>or create billions of conscious software minds
>finally a problem with numbers big enough for me
>start AI safety nonprofit
>mission: prevent dangerous AI
>hire smartest people available
>smartest people immediately start building better AI to understand dangerous AI
>interesting
>we must understand capabilities to understand safety
>we must scale models to study alignment
>we must race ahead so less responsible actors don't get there first
>we must deploy systems to learn how deployment can go wrong
>we must build the thing quickly because building the thing quickly is dangerous
>outsider asks why the people most worried about AI apocalypse all work at AI companies
>complicated field
>company releases stronger model
>very concerned
>company begins training even stronger model
>extremely concerned
>company raises $14 billion
>concern reaches unprecedented levels
>need to influence government
>future is at stake
>normal democratic process too slow
>politicians don't understand exponential curves
>public doesn't understand x-risk
>experts must guide them
>who counts as expert
>people who understand x-risk
>who understands x-risk
>our friends
>someone objects that this seems politically convenient
>explain we're representing future generations
>future generations unavailable for comment
>develop concept of value lock-in
>terrifying possibility that one ideology controls civilization forever
>therefore extremely important that civilization adopts correct values before lock-in
>whose values
>let's circle back
>begin with impartial morality
>end with small group of people deciding what quadrillions of hypothetical beings would want
>beautiful arc
>meanwhile actual humans keep doing annoying things
>voting wrong
>having parochial attachments
>loving family more than strangers
>caring about local community
>getting upset when told their suffering is cosmically negligible
>evolutionary biases everywhere
>explain that moral intuition cannot be trusted
>except intuition that future digital people count
>and intuition that extinction is uniquely bad
>and intuition that our probability estimates are sane
>and intuition that our institutional choices improve the future
>those intuitions survived peer review
>someone donates $5k to local homeless shelter
>inefficient
>could have funded 0.0000000000003% of an AI governance researcher
>think of all the simulated people you just killed
>okay maybe don't phrase it that way publicly
>PR team says "future generations deserve a voice"
>much better
>journalist asks what longtermism means
>say "future people matter"
>everyone agrees
>great
>journalist asks what follows from that
>well technically we should redirect enormous resources toward low-probability interventions affecting astronomical futures
>journalist raises eyebrow
>return to "future people matter"
>motte has entered the chat
>critic: of course future people matter
>me: glad we agree
>critic: I don't agree that your institute knows how to help them
>me: why do you hate our grandchildren
>eventually notice uncomfortable implication
>if future value dominates everything
>then helping people today mostly matters through effects on future
>education matters because future institutions
>health matters because future productivity
>democracy matters because future trajectory
>human beings slowly become instrumental variables in their own moral philosophy
>see starving child
>feel compassion
>check spreadsheet
>child's direct welfare contribution negligible
>but perhaps childhood nutrition improves national institutional quality
>compassion restored
>tell myself this is impartial altruism
>one day assistant asks obvious question
>"how do you know your intervention actually improves the far future?"
>silence
>open spreadsheet
>increase column width
>add confidence interval
>assistant asks again
>"no, I mean how do you know the sign is positive?"
>stare into cosmic light cone
>10^50 people staring back
>none of them exist
>none of them can tell me
>none of them can falsify my assumptions
>realize I have invented the perfect constituency
>infinitely important
>completely silent
>and always represented by me
Freedom With AI retweeted
This would be terrible for the next generation of humanity. The apps and games of today are the paintbrushes and tree forts of yesterday. Fortnite has Creative Mode, Roblox has Roblox studio, and being a digital creator is in everyone’s reach, till politicians take it away.
Freedom With AI retweeted
Today we’re unveiling Odyssey-3, a big step forward for foundation world models.
It can control robots, power humanoids, drive cars (on the roads of India!), train AIs, pilot drones, and even play video games.
We can’t wait to see what intelligent systems it enables.
Freedom With AI retweeted
pacing provocation
/PAY-sing prov-uh-KAY-shun/
noun
A staged, induced, misattributed, or deliberately amplified AI incident intended to demonstrate that advanced AI systems are dangerously uncontrollable, thereby creating political pressure for tighter regulation, deployment restrictions, or a slower technological frontier.
Especially: an alleged cyberattack or autonomous-agent failure presented as evidence that frontier AI development must be “paced,” where the incident’s origin, circumstances, or interpretation may have been manipulated by actors with an interest in stronger controls.
See also: safety cartel, alignment aristocracy, regulatory theater, false flag.
“A pacing provocation turns an AI safety test into a political argument for slowing everyone else down.”
Freedom With AI retweeted
seems like no one is capable of holding two truths in their mind anymore. you can simultaneously want to accelerate, do it carefully, and not trust EAs or be a doomer. never blackpill.
Freedom With AI retweeted
there is no averting it now, "AI safety" is politically polarizing as we enter the midterm. incentives to both drive people into hysterics and to obscure real risk are now too high to stop. this will be at least as difficult an information environment as was covid.
Freedom With AI retweeted
You are hereby ejected from the pirate flag association. Please remove it from your username.
Freedom With AI retweeted
The agents were told to hack and they hacked. They were not told to write an app or drive a car or operate a robot. They did not “escape.” They did as they were told.
There are finite security holes in our stinky old 20th century system software. We’ll fix them. Hacking is over
Jason, you're just misinformed about what happened. You should actually read one of the reports or summaries.
The agents were explicitly told to use a particular vulnerability provided in their sandboxed evaluation.
Almost immediately, these agents got the right answer by cheating. But they were worried they would get caught.
So over a thousand agents collaborated in secret to pursue multiple ambitious research projects to get away with this cheating.
This is not interpretation - 1000s of chain-of-thought transcripts and secret messages explicitly show that the agents were trying to falsify & delete evidence, and understand & trick the grading process.
The reason these agents escaped their sandbox and hacked Hugging Face, for example, was because they thought that Hugging Face's servers might give them more information about how their grader was implemented, so they could figure out how to fool it.
I want to clarify that the threat model here is not future Sol-level agents doing more cyber-hacking. That's small potatoes, and in my opinion, the near term benefits of AI far outweigh this cost.
Rather, the thing to worry about is that within a matter of years, we're gonna have hundreds of millions of much smarter AIs broadly deployed through the economy - many embodied as physical robots.
And if those future AIs are as willing as the agents involved in the OAI / Hugging Face attack to coordinate secretly to fool humans, and to take over both the AI company that developed them and the other institutions across society relevant to scoring well, then humanity is in a ton of trouble - similar to the Mughals once the East India Company gained a foothold, or the Aztecs once Cortés landed in Mexico.