well, they are. outsourcing is the bigger problem. just like outsourcing decimated manufacturing jobs, now it's decimating white-collar. i well and truly hope you get fired so you see how bad the market is. no other profession has this number of cruel, judgmental narcissists not looking out for their own kin.
> Also taking SSRIs near MDMA administration is one of the worst pieces of what is doubtless intended as harm reduction advice I’ve ever seen. You don’t hand people a recipe for serotonin syndrome because of a small scale rat test or two.
you have a deep understanding of releaser pharmacology if you state this. don't take it from me though--matthew baggott, one of the leading researchers studying enactogen pharmacology recommends doing this exact thing
dosed reasonably (~125 mg with no redoses), ideally at lower ambient temperatures, and ideally with an SSRI at the end of perceptible effects to stop MDMA to keep circulating through the neuron and causing oxidative stress, you will quite literally see 0 axonal damage. maybe some neurotransmitter depletion, but this is only temporary.
PCA yes that is concerning but I would be extremely surprised if it ended up in the MDMA supply chain. MDMA is made cheaply at quantities several magnitudes higher than research chemicals. it's much more likely PCA will be sold as a cut for others RC's -- especially as a part of those dreadful cathinone soups labelled "3-mmc" these days
and how is a black box algorithm any worse than a system where only large corporate gatekeepers control what you see? if anything, i'd argue tiktok is more fair and democratic.
it's interesting you bring that up as it turns out the incumbent party was actually the one behind that tiktok campaign, not russia. the campaign they used as evidence of "russian interference" that they cancelled the election because of
terms like "oversocialized" suggest you spend too much time on imageboards and you would do well to get off those sites. same with "humiliation ritual"
"One of the common solvents, ethyl acetate, comes naturally in many foods and beverages. It’s considered a safe chemical for decaffeination by the Food and Drug Administration."
speaking from a chemist's perspective, ethyl acetate is basically non-toxic. it decomposes into ethanol and acetic acid (vinegar). dcm (methylene chloride) is toxic, but should be completely removed before it ever reaches you.
the nice thing about volatile organic solvents is that they're...volatile, so they are easily stripped from coffee beans.
it does. that's what the "direct preference" part of DPO means. you just avoid training an explicit reward model on it like in rlhf and instead directly optimize for log probability of preferred vs dispreferred responses