Oh! Thanks for the heads up; we should use something else.
(If you have any suggestions, it’s been hard to find truly unladen symbols)
jayterwahl
Karma: 338
Special Persona Training: Hyperstition Progress Report 2
Silicon Morality Plays: The Hyperstition Progress Report
Re: elevated oral cancer risk in ALDH-deficient populations—I asked our dentech advisor Dr. Justin Merritt about it, and he said approximately,
The acetaldehyde cancer risk they describe is legit for AFR populations. However, connecting that risk to Lumina-derived ethanol production is where the argument becomes suspect.
Their argument is that a cariogenic diet (ie, sugar-rich) might produce sufficient ethanol in the mouth from Lumina to trigger local acetaldehyde production that damages the local epithelium; that the admittedly small total amounts of ethanol produced by Lumina might yield high local concentrations of acetaldehyde in the adjacent epithelium.
Firstly, S. mutans does not colonize the epithelium. It lives almost exclusively on enamel. The total surface area in the mouth that it could realistically inhabit is exceptionally small, unless Lumina can live in places that S. mutans generally does not. Other oral strep species overwhelmingly occupy the majority of epithelial surfaces of the mouth, which effectively restricts S. mutans to the teeth. This means that any miniscule quantities of ethanol produced by Lumina on enamel would be diluted by saliva before ever reaching the epithelium. Likewise, if other coinhabiting microbiota convert Lumina ethanol to acetaldehyde, this too would be diluted by saliva before reaching the epithelium.
Or, if one argues that Lumina could float around in saliva and affect oral epithelium by producing ethanol, I think this would be a dubious argument as well. Firstly, bacteria in saliva are transient, as they get swallowed. Secondly, the numbers just don’t add up in my mind. The saliva of people with extreme caries risk have 10^6 CFU/ml S. mutans detectable in saliva. At that concentration of bacteria (which is already quite rare), one could not even visibly detect turbidity in liquid. That shows how few S. mutans are present relative [to] the large volume of the mouth (i.e., it’s a pretty small quantity of ethanol production potential and even this level would only persist when sugar levels are high in saliva). S. mutans generally can only achieve high concentrations of bacteria in those specific places you tend to see caries develop, like the pit and fissures of the molars or interproximally.
The article makes a logical argument, but I suspect the actual risk is nowhere even close to what they propose. There are a number of estimates and guesses included in the article that may not pan out in reality. Fundamentally, the risks associated with Lumina-derived ethanol is miniscule compared to normal ethanol consumption. Is it possible that one who abstains from alcohol consumption could still be at risk due to Lumina ethanol production? I suppose it is possible, but I find it extremely unlikely.
Hi! I’m Aaron, Lantern Bioworks is my company.
Thanks. Similar to the concern of “could the bacteria colonize the vaginal tract after oral sex?” this is exactly the sort of edge-case risk we wanted to identify. We want to reward that safety-red-teaming by paying you $500. Send me your paypal, venmo, or crypto address at your leisure.
(The bug bounty is still $100 by default, but this was high-effort and we appreciate it. If anyone has more potential bugs to submit, email me at aaron@lanternbioworks.com !)
“Things we can’t talk about” in a relationship is another form of technical debt
I was a negative utilitarian for two weeks because of a math error
So I was like,
If the neuroscience of human hedonics is such that we experience pleasure at about a 1 valence and suffering at about a 2.5 valence,
And therefore an AI building a glorious transhuman utopia would get us to 1 gigapleasure, and an endless S-risk hellscape would get us to 2.5 gigapain,
And we don’t know what our future holds,
And, although the most likely AI outcome is still overwhelmingly “paperclips”,
If our odds are 1:1 between ending up in Friendship Is Optimal heaven versus UNSONG hell,
You should kill yourself (and everyone else) swiftly to avoid that EV-negative bet.
(noting the mistake is left as an exercise to the reader)
Where did you hear that TTS inspired dath ilan’s Governance?
Move Over, Insect Suffering: The Insect Vibing Hypothesis
I’m pretty bullish on “insect suffering” actually being the hedonic safe-haven for the planet’s moral portfolio
So as a K-selected species, our lives are pretty valuable, in terms of parental investment, time-to-reproductive-fruition, and how long we expect to live. As such, the neuroscience of human motivation is heavily tilted towards avoiding-harm; I think the studies say that people feel gains/losses at about +1/-2.5 valences; so, loss is felt much more sharply. (And maybe the average human life is hedonically net-negative for this reason; I go back and forth on that)
But for an R-selected species, we see all these so-reckless-they’re-suicidal behaviors. A fly is hellbent on landing on our food despite how huge and menacing we are. It really wants that food! A single opportunity for food is huge, in the fly’s expected lifespan, and if the well-fed fly can go breed, then it’s gonna pop out a thousand kids. Evolutionary jackpot!
But how much must the fly enjoy that food; and how little must it fear death, for us to see the behaviors we see?
I suspect the R-selected species are actually experiencing hedonically positive lives, and, serendipitously, outnumber us a bajillion to one.
Earth is a happy glowing ball of joyously screwing insects, and no sad apes can push that into the negative.
jayterwahl’s Shortform
Is the average human life experientially negative, such that buying three more years of existence for the planet is ethically net-negative?
(I predict that would help with AI safety, in that it would swiftly provide useful examples of reward hacking and misaligned incentives)
I imagine that WW3 would be an incredibly strong pressure, akin to WW2, which causes governments to finally sit up and take notice of AI.
And then spend several trillion dollars running Manhattan Project Two: Manhattan Harder, racing each other to be the first to get AI.
And then we die even faster, and instead of being converted into paperclips, we’re converted into tiny American/Chinese flags
I suspect that some people would be reassured by hearing bluntly, “Even though we’ve given up hope, we’re not giving up the fight.”
Not sarcastically! I wanted to have a Hard Mode available for those whose fasting was going well.
Vavilov et al certainly did it with seeds available.
I propose we surround ourselves in edible seeds, too.
Fair critique! Changed.
From my email exchange with Deadwyler, I took away that DARPA lost interest, and Deadwyler himself disappeared to go work for tobacco companies. And because orexin occurs naturally in the brain, it can’t be patented, which means that it’s hard to make money on it.
(I would expect a snortable cure for sleep would be worth something regardless, but I’m not a pharma company, so what do I know.)
Felt very different from modafinil; on a moda all-nighter I feel just mostly normal, kinda headachey, and a numbed fatigue. On orexin I still felt like I hadn’t slept; let me paste in my journal entry from that day (which is, to my great consternation, literally the only note which remains from this experiment).
“Feel spacey and out of it. More so than I did earlier, though that might just be that I’m interacting with people now.
Also hungry. More hungry than warranted? Not sure. Had a mealsquare around… 9? 10?
We ended up using the string “XXF”