Scholarly consensus is that the "Gospel of Matthew" was not written by the apostle Matthew and the "Gospel of John" was not written by the apostle John:
For that assertion to hold water, "scholarly consensus" would have to define "scholarly" so narrowly as to exclude the vast majority of scholars (it seems like it should go without saying that most scholars in this area are Christian who maintain apostolic authorship).
Perhaps they are dismissing scholars who identify as Christian? That would be quite the catch-22.
To me, it is apparent that the data cannot support any clean division between two "sides", it tells a more complicated story about sometimes there was apostolic authorship, sometimes not, and sometimes we don't really know.
I would suggest that the real academic consensus is that we can confidently rule out the us-vs-them preoccupation that is common in lay discussion.
"No sides in science" is a silly idea. Of course, scholars have biases. They're human. Humans like to group up and gang up against other.
Specific to Bible Scholarship, I wager the two big sides are scholars who have faith (i.e., Nicene Creed) and scholars who have little. Bruce Metzger who had some faith, and Bart Ehrman who has none. RSV/ESV which says Jesus is the "Son of God" in Mark 1, and NRSVue which deletes "Son of God" from Mark 1.
There are plenty of YouTube videos that go into the subject thoroughly. I couldn't find the one I watched recently stating the notion that the gospels ever could have been totally anonymous is absurd. Nobody would take you seriously, reputation was everything in the ancient world. The people of the time knew exactly who wrote what, even if there weren't any direct titles on the actual manuscripts.
So then who wrote Hebrews? It wasn't Paul's writing style, and it doesn't name it's author. Matthew and Luke don't name the Q source material they have in common. Let's take gMark, someone composes it around 70AD somewhere. It gets copied and sent to other communities elsewhere. Decades later it's attributed to Mark.
Reputation has never been everything & as crazy conspiracy theories like Qanon & antivax prove, some sizable fraction of the population will find a way to believe whatever they want to.
Sure. They had crazy conspiracy theories back then too. Anyone can believe what they want. But reputation means something today just as it did back then. It's just today we outsource that function to the academic system.
Where do the attributions come from, Papias? He claimed Mark wrote down Peter's teachings in the wrong order, and that Matthew's gospel was written in Hebrew. But the Matthew we have is in Greek, copies from Mark and shares other Greek material with Luke (Q source).
They're not proposing to apply tensor decomposition to an existing collection of weights. It's an architecture in which the K, V, and Q tensors are constructed as a product of factors. The model works with the factors directly and you just need to compute their product on the forward pass (and adjoints on the backwards pass), so there's no decomposition.
You're totally right there must be supervision; it's just a matter of how the term is used.
"Supervised learning" for LLMs generally means the system sees a full response (eg from a human expert) as supervision.
Reinforcement learning is a much weaker signal: the system has the freedom to construct its own response / reasoning, and only gets feedback at the end whether it was correct. This is a much harder task, especially if you start with a weak model. RL training can potentially struggle in the dark for an exponentially long period before stumbling on any reward at all, which is why you'd often start with a supervised learning phase to at least get the model in the right neighborhood.
Backprop itself doesn't invert the computation, but it does give you the direction for an incremental move towards the inverse (a 'nudge' as the article puts it). That is, given a sufficiently nice function f and an appropriate loss ||f(x) - y*||^2, gradient descent wrt x will indeed recover the inverse x* = f^{-1}(y*) since that is what minimizes the loss. I assume this what the article is pointing at.
If you want to be picky, it's true that the direct analogue of continuous optimization would be discrete optimization (integer programming, TSP, etc) rather than decision problems like SAT. But there are straightforward reductions between the two so it's common to speak of optimization problems as being in P or NP even though that's not entirely accurate.
Nitpicking, but for a technical audience it's worth noting that ibogaine is not at all a 'potent' psychedelic in the pharmacological sense of the term. A typical therapeutic dose is on the order of 500mg, which makes ibogaine something like 20 times less potent than psilocybin (typical dose ~25mg), which itself is 100 times less potent than LSD (typical doses less than 250ug).
Of course, this isn't really relevant to the subjective experience of taking ibogaine at its typical dose, which by all accounts is strange in ways that go beyond the classical psychedelics.
sort of an interesting dichotomy in the meaning of potency. The poison that kills you with the smallest dose is the most potent, but the (hypothetical) drug that cures cancer most effectively is the most potent cancer cure, regardless of dose.
One definition of potency is how little mass it takes to do its thing, whatever that is. Fentanyl is a potent drug, LSD is a potent psychedelic. That resolves the dichotomy.
There’s a second, softer connotation, of how strong it is in light of whatever other limits may exist. Mescaline is less potent in this way, not just because of the larger mass, but because a dose that’s going to blast you into another realm is going to be much harder on you than some other psychedelics.
Ibogaine (haven’t done it) is very potent by this meaning. Very long, intense trips are possible.
all the poisons kill you, one of the poisons is most potent
only one of the cancer cures cures you, but you have to drink a swimming pool of it; thankfully, it's potent enough in that dose to kill the cancer where the other choices only slow it down
> Of course, this isn't really relevant to the subjective experience of taking ibogaine at its typical dose, which by all accounts is strange in ways that go beyond the classical psychedelics.
> Altman was initially going to cooperate and even offered to help, until Brian Chesky & Ron Conway riled him up
I don't think the article supports this. All we know is that sama appeared cooperative when the board fired him. This was probably a reasonable posture for him to adopt regardless of his actual intentions at the time.
I believe it. Note that this story is being sourced from Altman/Conway, even to the level of their private text messages. So they would have to be the ones fabricating this claim, but this story is embarrassing to them: if they were going to make it up, Altman's change of heart would be prompted by appeals from employees or the board (which was in fact the version of the story that was initially circulating on social media and Altman is still trying to spin as the reason the Board eventually called him). As it is, it comes off as duplicitous & destructive and highly unflattering: 2 rich CEOs/VCs riling him up to go back on his promise to try to take over and burn down OA if he can't.
Counterpoint: SSRIs were transformative for my depression. Side effects were minor and manageable (eg, Wellbutrin worked well to prevent any sexual dysfunction). I was on them for several years and had no problem tapering off. My understanding is that this is a pretty typical experience. Rhetoric like this was actively harmful in dissuading me for years from trying what ended up being by far the most effective treatment for me.
(yes, I've tried psychedelics; they're fascinating and super promising, but at least for me, not transformative in the way that fluoxetine was)
No individual depression treatment works for everyone. SSRIs are not a magic bullet. Neither are psychedelics. But if you're depressed and haven't tried SSRIs, you owe it to yourself and everyone in your life to at least test the hypothesis that they might help.
Scott Alexander's page on SSRIs is a great, relatively objective resource, from a psychiatrist who regularly prescribes them:
https://lorienpsych.com/2020/10/25/ssris/
Same. I'm a big time psychedelic advocate, and relatively frequent consumer.
Prozac actually saved my life. Psychedelics since have certainly enhanced it, but they could not accomplish what prozac did. I can go into more detail if interested.
Please do, I’m genuinely interested in your experience because, in my experience (!!) with family and acquaintances, I’ve only ever see Prozac as a part of a turn for the worse. I do acknowledge I might not be aware of cases where it was neutral or positive, however.
What was your experience like? And do you still take it?
> I’ve only ever see Prozac as a part of a turn for the worse.
You might be interested in the book "Empire of Pain: The Secret History of the Sackler Dynasty"[1] by Patrick Radden Keefe. Although it focuses on Oxycodone and other painkillers it touch on Prozac and other SSRI's. Quite the eye opener...
Ditto with family. One antidepressant led to another, and then to another, and on to deeper depression, suicide attempts and more. Finally landed through the prescription pill grapevine on Xanax, which by far, without any comparison whatsoever, was the most persistently destructive and frightening. Like dropping a nuclear bomb on the whole family, for years and years.
I don't mean to bring other classes of drugs in, only to note that it is _very common_ for someone to go in for one thing, find it doesn't work, and then to start tweaking / adding / combining in an attempt to find a solution, and that's where the real problems often live.
When mental health is so bad it messes with work and personal relations one may feel obligated to share. When things start going well there's no need to explain what's going on. I've seen many that it works for and some where they just had to keep looking at other options.
Fully agree SSRI:s probably saved my life. SSRI:s are supposed to be used as part of therapeutic regimen with therapy and regular checkups with doctor, not as over-the-counter remedy like aspirin. I’m guessing that using them detached from any therapeutic context increases the prevalence of negative experiences.
That certainly could be a factor in some people having bad experiences in America at least–as I understand it insurance coverage for mental health is often limited or bad, and appointments with doctors are hard to come by and often only last 15 minutes.
Well the issue is finding a therapist you like that is also covered. I just moved and have been looking for a new therapist. All the ones that I’ve been recommended are not part of an insurance network. You can file for reimbursement, but it’s slower and the coverage isn’t as good as in network doctors. I went from a $20 copay for my last therapist to paying $200-300 per session out of pocket (depending on who I go with).
Psychiatrists can have similar issues but you don’t meet with them nearly as often. Once I got on meds that worked for me, we’d have a 30 minute meeting every two months just to check in and see if any adjustments were needed. That was fine by me because I was going to a therapist weekly, the two were aware of each other and would communicate if needed, and it saved both of us time.
I think in my part my negative reaction is, like many people, having observed the effects of SSRIs on young people. It's well known that risks like suicidal ideation are actually higher among those under 25, and in general it is awful to see the mental health crisis among young people dealt with primarily via instantly reaching toward semi-permanent medication, rather than considering other treatments.
I don't know if reacting to current events to offer medical advice is the best way to go. SSRIs were of great help to me when I was in the throes of my anxiety disorder, but psychadelics had given me harrowingly bad trips before I started a meditation practice (the mental fortitude has really helped.) One particular trip made me feel like I just came back from hell and I didn't feel "back" for days afterward. Mental health is tough and nobody seems to have the answers. As with everything, following the beaten path will lead you to surfaces well-trodden, and answers at this point in time are more personal than known science.
The most reasonable conculsion (though tentative) seems to be that SSRI and the more "traditional" option may help some, while psychedelics may help others. (And we have to imagine other drugs in the future will be found.) The trick will to figure out which person will respond to which. But I think all can agree having more options is better.
Yes. My root comment is not giving medical advice, but stating the unspoken truth in the original article that people are reaching for psychedelic therapy because existing standard medications do not work for everyone. Real medical advice should be tailored to an individual's history and needs, which is what any good physician or psychiatrist would actually do.
I take Wellbutrin and I’m not sure what it does for other people, but for me, it gives me a buffer of patience to draw from that I didn’t really have before. Negative or frustrating events don’t feel quite as urgent and I’m able to calmly move forward better than I can without it. Which is very helpful when you’re taking care of kids.
This is my experience, too. Bupropion (Wellbutrin) seems to dampen the spike of cortisol or adrenaline or something that I can get when something really irritates me. As a single father of two middle school age boys, it has helped a lot.
Btw, have you noticed exposure to sunlight to make you even more laid back? I've recently been casually tracking a correlation in that for myself.
This has been my experience with citalopram as well. The negative side effect though is that natural highs are lower, so I view it very much as a mood stabilizer.
A lot of medical treatment is like this. Everything has side effects, but at least with regulated, approved treatments, they're well-known side effects with a large-sampled quantified probability profile, thanks to decades of use following years of clinical trials. If you just heed the Internet for anecdotes, though, all you ever hear are the horror stories and your personal risk assessment becomes biased away from statistics and toward compelling stories.
For what it's worth, synthetic opioids and spinal fusion saved my life. If I'd listened to the Internet, I'd have likely never pursued treatment or maybe just taken Kratom and gotten massages or something, fearing I'd end up a drug addict with a worse spine than I started with.
How did you decide to stop? I’ve been on Escitalopram for a few months now after suffering depression and anxiety that affected (and was affected by) my work relationship. My GP tried to get me down from 10mg to 5mg but I felt terrible again. I’m now on 7.5mg and feel ok. But I can’t imagine stopping. Like ever. I wonder how I could stop. Or perhaps just accept and take them forever?
I was on escitalopram for a while. Then I added Wellbutrin, took away the escitalopram, added fluoxetine, and now I’m off all of it.
My advice is to take a breath and focus on the present. Chip away at what’s causing your anxiety. Don’t shy away from working on yourself. Use your medication as a tool, not as a solution.
I’m painfully aware that these are all much easier said than done. It was a three year journey for me. And who knows - I may start having panic attacks again. If I do, I’m going to try buspar.
Also. Don’t be afraid to switch medications. Don’t just live with severe side effects. Sexual dysfunction, lethargy, significant weight gain - this is no way to live your life. I found the Prozac - Wellbutrin combo to be pretty tame in that department.
Prozac is my fun drug. It makes me hypo manic with one 20mg dose.
I have schizoaffective disorder, bipolar type So I am prone to being more sensitive to SSRI. This is mostly to do with my genetically odd 5HT2A receptors.
Depression is no longer really a part of my life anymore after I found I was zinc deficient. But now I do tend towards the manic and have issues with psychosis still so I need to be careful with my serotonin.
How do you get the right amount of zinc? Are there other minerals you're supposed to take with zinc? I know that sometimes balancing one thing causes another thing to go out of balance.
Same for me. It's hard to say exactly what SSRI's are doing for me but I definitely feel better.
The bad sides of of SSRI's are as overblown as the good sides of psychedelics are. It's easy to form an opinion from reading personal experiences online but that doesn't reflect the real world imho.
I don't think Scott Alexander's page is a 'relatively objective resource'. He is, as you point out, a 'psychiatrist who regularly prescribes them' and also used to take them. It is full of special pleading where he relitigates extremely high quality meta-analyses to overlay his own opinion:
> That is, there are a bunch of tests that ask you a bunch of questions about your feelings and symptoms, and you can add them up and call that a “depression score”, and if you do that, antidepressants have an effect size of 0.3. Or you can ask patients “how depressed do you feel on a scale of 1-10”, and if you do that, antidepressants have an effect size of 0.5. I think the latter is better, because it’s what we actually care about (how patients are doing), and the tests are kind of dumb and ask about a lot of symptoms most people realistically aren’t experiencing.
(In other words, if you ask a patient with depression how they are feeling, and they say 'great', and then you ask them questions like "are you managing to shower every day", or "did you think about suicide a lot this week" and they give the same answers as a depressed person, they are cured!)
Does weird napkin math which clearly can't be justified:
> For those people, they will have a large real effect size of 1.0, plus a large placebo effect size of 0.9, for a very large total effect size of 1.9.
(How do you get to add the placebo effect back on to the postulated 'large real effect size'??)
Says that extremely common side effects are 'very unusual':
> It can be any or all of decreased libido, difficulty orgasming, difficulty getting an erection, difficulty enjoying sex, or decreased sensation in the genitals. These usually go away a few weeks to months after stopping the medication, but in rare cases they might linger for months or years, and there are a few people who say their sexual side effects never went away. These cases are very unusual and still not well understood.
(Note that in the same article he points out that, in general, the medication only improves mood or anxiety while you keep taking it, when you stop taking it you still have the depression or other conditions. So the fact that sexual disfunction usually gets resolved after stopping taking the medication isn't much relief. For most people SSRIs will never lead to a steady state where you are stable with regards to your mental health issue and also are able to enjoy sex.)
Makes armchair psych connection between well-studied things which are not the same:
> When everything goes right, SSRIs blunt negative but not positive emotions. But many people even at reasonable doses will notice that their most extreme positive emotions become a little less extreme (this may be part of the problem with sex).
(Difficulty getting aroused or orgasming or feeling in the mood for sex is not the same as "most extreme positive emotions becoming a little less extreme")
Personal anecdatum, different SSRIs affect people differently. Some that others here have praised didn't help, or had bad side effects.
What I am on now (Lexapro) was life changing in a good way, with only minor sexual side effects that more or less went away.
As such, studies that day "SSRIs have this effect on people" or "have these side effects" are fundamentally flawed. Despite belonging to a common class, there isn't a universal experience.
Ok, so the middle-brow infotainment psych blogger's essay is valid, but meta-analyses are 'fundamentally flawed' because 'different SSRIs affect people differently'?
This news will probably come as a huge surprise to the psychiatric epidemiologists who carried out the peer-reviewed research that Alexander mangles! They probably thought all SSRIs were exactly the same!
I was with you up to this. What even is a “psychiatric epidemiologist” I had to look it up,
“It is a subfield of the more general epidemiology. It has roots in sociological studies of the early 20th century. However, while sociological exposures are still widely studied in psychiatric epidemiology, the field has since expanded to the study of a wide area of environmental risk factors, such as major life events,“
Yeah, I don’t know, I’d sort of go with an experienced clinician when it comes to advice about pills. Sociology/epidemiology is cool, but there’s a lot to say for the importance of “practice” in medicine.
Good rebuttal. The person who did the peer-reviewed statistical research is wrong because Wikipedia describes their discipline as related to sociology.
No, I’m just saying that for me personally, if I wanted advice about taking psychiatric medication, I would prefer a clinicians view over the view of a epidemiological researcher, _even if they have the exact same level of training as medical doctors_. I don’t know about the researchers in question, maybe they too are practitioners. If they are, you should qualify that they aren’t merely epidemiologists. Epidemiology as a field has a problem with replication not unlike psychology or nutrition research.
A meta-analysis of SSRIs isn't very useful, since it draws conclusions about a group of drugs that do not have similar behaviors within an individual.
It's a bit like doing a meta analysis of hydrocarbons in two and four stroke engines. Some will work better than others in some situations, but the meta analysis itself isn't illuminating when you are putting liquid propane into a gunked up carburetor.
Wellbutrin also excarcbates OCD symptoms with typical responses including paranoia, hallucinations, eventually inducing psychosis. The article you post the author admits they don't really understand how SSRIs work. Why should putting such things in your body be any different than psychedelics or heroin or eating chocolate when you are sad? Simply put, it's not. This kind of science is based on collecting the minimum number of people to establish a p value and effect size larger than 0. You do that enough times then you can give your pills to whoever you want. Psychiatry provides a physiological change, i.e. they give you pills, based on a diagnostic criteria that removes all physiological reasons for the presented symptoms. This all seems rather backwards to me. Especially when modern psychology has developed therapies like cognitive behaviour therapy, schema therapy, dialectical behaviour therapy, which all basically can be summed up to state "your thoughts and feelings don't matter and probably get in your way. What matters is how you organise your mind and actions." So modern psychology has arrived at stoicism as the answer to a variety of problems such as depression, post traumatic stress disorder, obsessive compulsive disorder, and personality disorders. This isn't to say that pills don't help. It's just that they are not a panacea nor are they well understood by the psychiatrists who are motivated to give you pills since their training as a medical doctor teaches them peoples problems have physiological solutions.
Prophet has gotten a lot of attention since being released in 2017, I think because the idea of a fully automatic solution is very appealing to people. One of the original developers, Sean Taylor, recently posted a nice retrospective on the project's successes and failures:
https://medium.com/@seanjtaylor/a-personal-retrospective-on-... He quotes one of his earlier tweets:
If I could build it again, I’d start with automating the evaluation of forecasts. It’s silly to build models if you’re not willing to commit to an evaluation procedure. I’d also probably remove most of the automation of the modeling. People should explicitly make these choices.
Having worked on similar Bayesian time-series forecasting tools at Google, this matches my experience (though I've never used Prophet seriously, so please don't take this as any direct judgement of it as a software package). There is a lot of value in a framework that lets you easily experiment with different model structures (our version of this was the structural time series tools in TensorFlow Probability, see, e.g., https://blog.tensorflow.org/2019/03/structural-time-series-m...). But if you're forecasting something you actually care about, it's usually worth the time to try to understand yourself what structure makes sense for your problem, and do a careful evaluation on held-out data with respect to whatever metric you're really trying to optimize. A fully automated search over model structures is cute, but even when it works, it mostly just ends up rediscovering properties of the data you could or should have already known (e.g., of course traffic to your work-related website will have a day-of-week effect), so the cases where it really adds practical value are harder to find than you might like.
Even in the age of deep learning, I do think these relatively classical Bayesian models have a lot of value for many applications. Time-series forecasting tends to be a case where:
- you don't have a ton of iid data points (often, only a single time series),
- you'd like forecasts with principled uncertainty estimates, e.g., credible intervals, giving you a range of scenarios to plan for,
- you often do have a pretty good idea of what features are relevant to the process you're predicting, and
- you want to understand in detail what features the forecast is accounting for (and what it might be missing),
all of which play to the strengths of more classical, structured statistical models, compared to more data-hungry black-box deep learning models. So the basic ideas in Prophet and similar tools do still have a lot of relevance going forward, IMHO.
You mention classical models but Bayesian deep learning is a thing too. One can even retrofit existing DL models to obtain uncertainty estimates, at the expense of increasing (possibly doubling) the number of model parameters.
The quality of the uncertainty estimates is a question though.
Before trying MDMA I had mostly written it off as a feel-good drug: an artificial high like cocaine or meth, useful only for escapism. Why bother chasing that sort of experience? But now having tried most of the commonly used psychedelics, I've come to believe that MDMA is the most profound of them all.
MDMA does feel good, of course, but it's not escapist. It’s a deep, wholesome, fundamentally healing sort of goodness. It is unconditional, redeeming love and forgiveness — the core of Christian spirituality. It is the revelation that you really are lovable, even your darkest, hidden parts, and that you are capable of love. Debatably, there is no more profound lesson to be learned about the human condition. It really is magical.
Even so, the experience is surprisingly subtle. It doesn't particularly force positive feelings ("ecstasy" is a total misnomer, IMHO). At first you don't necessarily even notice any effect at all, maybe just a mildly better-than-average mood. But gradually it becomes clear that this subtle sense of well-being is infinitely deep: nothing you might experience can possibly disturb it. All sense of shame and self-judgement, fear of rejection, hang-ups that get in the way of connecting with people --- all dissolve immediately on contact. And from that sense of absolute safety, the capacity to love emerges naturally. The drug doesn't generate it. It just helps you get out of your own way.
I've taken MDMA a few times now just on my own at home (lacking a rave community, although I'm sure that's a fantastic experience also), where it's relatively easy to implement harm reduction measures: stay hydrated, take protective supplements, get a full night's sleep before and after, and wait multiple months between doses (doing all these, I've never experienced a ‘hangover’, just a positive afterglow). I've found the most rewarding results from trying to keep my attention grounded in bodily sensation, gently returning to the body whenever I notice I've become lost in thought. Often, difficult memories or associations will surface of their own accord, sensing that it's safe to do so, and seeing them from a loving perspective can be immensely healing.
I really hope we can eventually find our way to making this experience legally and safely available to everyone who wants it. Yes, MDMA has sharp edges; it's not as physiologically benign as the classic psychedelics, but it's not addictive and it can be used safely. Not everyone has good experiences every time, but compared to the classical psychedelics, it's much more reliably positive. It apparently has some effectiveness as a medicine for specific illnesses like PTSD, but IMHO the real condition it treats is much broader: the universal human condition of feeling more walled off than we'd like to be.
One last galaxy-brain thought: if we ever figure out a way to replicate MDMA's pro-social effects that people could safely use on a day-to-day basis, it might be the most valuable thing we ever invent. One could even see it as the metaphorical second coming of Jesus, his kingdom on earth achieved through purely secular means. How's that for a career goal? :-)
For what it's worth, Jax (github.com/google/jax) now lets you use XLA to compile Numpy code. It'd be cool to see how that would stack up in a modern comparison.
https://en.wikipedia.org/wiki/Gospel_of_Matthew#Author_and_d...
https://en.wikipedia.org/wiki/Gospel_of_John#Authorship