Posts
Comments
Given that the model eventually outputs the next token, shouldn't the final embedding matrix be exactly your linear fit matrix multiplied by the probability of each state to output a given token? Could you use that?
This is extremely cool! Can you go into more detail about the step used to project the 64 dimensional residual stream to 3 dimensional space? Did you do a linear fit over a few test points and then used it on all the others?
I think you could, but then it would be unintelligible to most people who don't know wtf is Solomonoff Induction.
The Ponzi Pyramid scheme IMO is sn excellent framework, but the post still suffers from a certain, eh, lack of conciseness. I think you could make the point a lot more simply with just a few exchanges from the first section and anyone worth their salt will absolutely get the spirit of the point.
I think this is an added layer though - I don't think the responses listed here are responses of people deep enough in the transhumanism/AI rabbit hole to even consider those options. Rather, they sound like the more general kind of answers that you'd hear also in response to a theoretical offer of immortality that means 100% what you expect it to, no catches.
If immortality becomes widely available, we will lose the current guarantee that "awful people will eventually die", which greatly increases the upper bounds of the awfulness they can spread
I mean... amazingly good people die too. Sure, a society of immortals would obviously very weird, and possibly quite static, but I don't see how eventual random death is some kind of saving grace here. Awful people die and new ones are born anyway.
I think another big issue with codes of conduct is that they just shift the burden around. You're still left with the issue of interpreting the spirit of the norm, deciding if everyone at least made a good faith attempt to stick to it, if good faith is enough, etc. I don't have much experience with them but I honestly don't know if they help that much. Seems to me like there are two types of "troublemakers" in communities no matter what:
- people who are purposefully deceptive and manipulative;
- people who simply lack the social grace and ability to "read the room" required to meet other's expectations of social norms adherence rather than just stick to their own interpretation of them.
Type 1 you want to kick out. Type 2 you ideally want to be a lot more graceful and forgiving with, though in some extreme cases you might still need to kick them out if their problems are unfixable and they make no effort whatsoever to at least mitigate the issues. Writing the rules down doesn't help as long as they're flexible, because the problem those people have is a lack of the sort of intuition that others possess for grokking flexible rules altogether. And if you make them inflexible you just have a chilling effect on every interaction, and throw away a lot of good with the bad. After all, for example, why shouldn't someone ask a woman out at their first meeting if they're both clearly into each other and sparks are flying? These things happen! And people should be able to give it a try, I think it's important to make it clear that there's nothing sinful or bad about courtship or flirting per se; too many rigid rules about such personal interactions inevitably carry a sort of puritanical vibe with them, regardless of intention. But as usual, "use your best judgement" has very uneven effects because some people's best judgement is just not that great to begin with, often through no fault of their own.
I would. It's possible an election in which a third party candidate has a serious chance might exist, but it wouldn't look like this one at this point. Only way the boat could at least be rocked is if the charges go through and Trump is out of the race by force majeure, at which point there's quite a bit of chaos.
I mean, even so... it's ten minutes. I'd be bored on a 2 hour trip on which I'm unable to read. For ten minutes, I can manage.
Shared biological needs aren't a guarantee of friendliness, but they do restrict the space of possibilities significantly - enough, IMO, to make the hopes of peaceful contact not entirely moot. Also here it comes with more constraints. Again, if we ever meet aliens, it will probably have to be social organisms like us, who were able to coordinate and cooperate like us, and thus can be probably reasoned with somehow. Note that we can coexist with bears and chimpanzees. We just need to not be really fucking stupid about it. Bears aren't going to be all friendly with us, but that doesn't mean they just kill for kicks or have no sense of self-preservation. The communication barrier is a huge issue too. If you could tell the bear "don't eat me and I can bring you tastier food" I bet things might smooth out.
AI is not subject to those constraints. "Being optimised to produce human-like text" is a property of LLMs specifically, not all AI, and even then, its mapping to "being human-like" is mostly superficial; they still can fail in weird alien ways. But I also don't expect AGI to just be a souped up LLM. I expect it to contain some core long term reasoning/strategizing RL model more akin to AlphaGo than to GPT-4, and that can be far more alien.
This is a double edged sword to me. Biological entities might be very different in the details but shaped by similar needs at their core - nutrition, fear of death, need for sociality and reproduction (I don't expect any non-social aliens to ever become space faring in a meaningful way). AIs can ape the details but lack all those pressures at their core - especially those of prosociality. That's why they might end up potentially more hostile than any alien.
The notion that planetary spread will cause necessarily war is IMO hugely flawed because it ignores entirely the issue of logistics. People don't make war just because they piss each other off - I mean, sometimes they do, but war also has to be at least practical. Logistics of interplanetary or, heavens forbid, interstellar war are beyond nightmarish, which is why space operas always come up with jump drives or wormholes or gates or some other kind of technoblabbery doohickey to make wars across galactic empires work much like wars on this little mudball we're used to. Otherwise, the universe has a very very strong "live and let live" bias; plenty of real estate, plenty of buffer zones in between, and it's almost always cheaper to go somewhere empty than to wrestle somewhere full from the hands of someone else, especially if you want the planet to stay intact and livable.
There are precedents on Earth too. The Roman Empire and early Qin were both very powerful, very large, and very expansionistic, separated by thousands of years of cultural and technological divergence. According to this theory, they should have been natural enemies who went to war almost immediately. And yet they didn't, first and foremost because in between them was a lot of inhospitable land that neither side could economically cross without arriving to the other substantially weakened. And also because they were probably different enough that they didn't really concern themselves with mutual annihilation on ideological basis - that's more the province of the devil you know, the heretic, the guy who's similar enough that you care but different enough that he pisses you off. You don't fight a Thirty Years' War with some distant off culture that believes completely different things, you fight it with your brothers and sisters who dared believing a slightly different version of what you believe (and then, hopefully, you learn not to fight it at all because it's really self-destructive and stupid).
Obviously there are risks - it's true that space colonies would diverge from Earth for sure, and it's true that having humanity spread on multiple planets would make the use of even potentially planet-ending weapons like nukes or relativistic kinetic bombardment a bit less taboo. That's a problem, but it does not mean that History in such a future would be more predetermined than it ever has been.
That is an interesting counterpoint, but there's the fact that things like PRISM can exist in at least something like a pseudo-legal space; if government spooks come to you and ask you to do X and Y because terrorism, and it sounds legit, that's probably a strong coordination mechanism. It still came out eventually.
To compare with COVID-19, there probably are forms of more or less convergent behaviours that produce a conspiracy like appearance, but no space for real large conspiracies of that sort I can think of. My most nigh-conspiratorial C19 opinions are that early "masks are useless" recommendations were more of a ploy to protect PPE stocks than genuine advice, and that regardless of its truth, a lab leak was discounted way too quickly and too thoroughly for political reasons. Both these though don't require large active conspiracies, but simply convergent interests and biases across specific groups of people.
Yeah, that's what I meant when I said people can't do that sort of thing on that scale without screwing it up. It just breaks down at some point.
Training: it's not clear to me whether training LLMs on copyrighted content is a copyright infringement under the current US copyright law. I think lawmakers should introduce regulations to make it an infringement, but I wouldn't think the courts should consider it to be an infringement under the current laws (although I might not be familiar with all relevant case law).
My feel is that it could have been fair use as long as LLMs were just research projects, but then OpenAI started selling theirs as a product without changing their working model at all, and if you're commercializing the model, it's another story. Not sure where open models would lay here, but I still reckon probably copyright infringement since you're not using the data only internally. Would like to hear an expert's opinion on this.
I think to the extent LLMs don't preserve the wording/the creative structure, copyright doesn't provide protection; and some preservation of the structure might be fair use.
The problem is that this is a new case because it completely destroys the business model of these websites if you can have an AI agent visit them and then relay a summary to you - as it denies them the clicks (and ad visualizations) they need to pay themselves off. At which point odds are they'll just lay on paywalls even harder if they're not protected from this.
ChatGPT hallucinating false info and attributing it to NYT is outside copyright law, but seems bad and damaging.
I could imagine something like a defamation lawsuit? But it would probably have to focus on a specific case, not the general possibility of it? Again, hard to guess, this is all unexplored territory and new questions that never needed to be asked until now.
Man, I'm so crossing my fingers that Betelgeuse kicks the bucket before I do.
For a long time western philosophy literally believed the stars to be encased in a single rigid sphere surrounding the Earth. In fact it was a really big deal in Aristotelic cosmology that the heavy, corruptible, changing things are low (and thus fall to the Earth) while the substance of the skies is higher and incorruptible, thus unchangeable. It's why Copernican and Newtonian celestial mechanics were such a big deal. The mind-blowing part was that they suggested that everything everywhere followed the same laws, and the sky wasn't special in any way.
That said, this is just the view that had become mainstream in medieval Europe. If you had asked Democritus back in ancient Greece, he'd likely have told you that the stars were just other suns like ours, with other planets like ours, moving through the void, because that was the atomist view.
To be fair, the star patterns do have vague shapes that inspire their name (in Orion the shape of a man with a club and a shield is quite recognizable). But of course pareidolia was applied out of necessity as you say; it was mnemonics probably as much as a good bit of storytelling fun. I would also add another possible mechanism: passing through multiple cultures. Maybe you inherited the constellation name from a previous culture, and mapped their God or hero of choice with one of yours because that makes it more familiar. But maybe some part of the analogy was lost in the mapping. Do it enough times and name and constellation may seem now associated in a completely arbitrary way.
Man, I have to wonder around what fire did a human storyteller first go "wtf am I saying, I know tradition and all, but that's six stars, not seven" and decide to make up an excuse for it.
In my experience, the problem with arguments against COVID-19 vaccines is that they mainly consist of evidence that there's risk involved in getting vaccinated. To usefully argue against getting vaccinated, one needs evidence not only that vaccine risks exist, but that they're worse risks than those of remaining unvaccinated.
In the most extreme cases they also assume vast conspiracies that I assign a very low prior probability to simply based on "people can't do that sort of thing on that scale without screwing up". Paradoxically enough, very often conspiratorial beliefs assume more rationality and convergent behaviour (from the elites seen as hostile) than is warranted!
I think some of these cases aren't irrationality at all, but difference in partial information. Belief trajectories are not path-independent: even just learning the same events in different order will lead you to updating differently. But there's also irrationality involved (e.g. I'm very unconvinced about arguments against the gambler's fallacy when there are people who express it in so many words all the time), it just shouldn't be your go to explanation to dismiss everyone else but yourself.
No risk is zero, that's not a reasonable way to think about control over one's life. And you don't choose partners at random so intelligence send conscientiousness in couples probably correlate far better than that.
I mean, .01% is a tiny rate if we're talking yearly, and I think an acceptable risk even if you don't plan to abort. But at this point we're completely away from the original point because .01% would mean you have a lot of control, which is my point exactly. Nothing is perfect, but the estimated efficacy of good contraception is probably largely dragged down by a long tail of people who are really bad at it or downright lie in self reporting studies.
You mentioned things like women who lie about contraception and split second decisions, which IMO are nonsense to bring up in this context. But going back to condoms: yes, I believe that 3% figure to be garbage. The 3% figure is average and based on people self-reporting. But in practice, condoms are hard to break, and even if they do break it's easy to realise. Morning after pills are a thing for "accidents" you notice. So IMO reasonably conscientious people that actually use condoms properly (rather than just saying so in questionnaires) and double down with morning after pill in case of accidents will achieve a much better rate. 3% is an upper bound, because it includes a lot of confounders that skew the rate to be worse.
I mean this goes into the philosophical problem of whether it makes sense to compare utility of existent and virtual, non-existent agents but that would get long.
If your timelines are short-ish, you could likely have a child afterwards, because even if you're a bit on the old side, hey, what, you don't expect the ASI to find ways to improve health and fertility later in life?
I think the most important scenario to balance against is "nothing happens", which is where you get shafted if you wait too long to have a child.
There are biological problems you might not know you have, there are women who lie about contraception, there are hormonal pressures you won't feel till you reach a certain age, there are twins and stillbirth, and most of all there are super horny split second decisions in the literal heat of the moment that your system 2 is too slow to stop.
This is absolutely nonsense IMO for any couple of grown ups of at least average intelligence who trust each other. People plan children all the time and are often successful; with a little knowledge and foresight I don't think the risk of having unplanned children is very high.
I think the idea here was sort of "if the kid is unaware and death comes suddenly and swiftly they at least got a few years of life out of it"... cold as it sounds. But anyway this also assume the EY kind of FOOM scenario rather than one of the many others in which people are around, and the world just gets shittier and shittier.
It's a pretty difficult topic to grasp with, especially given how much regret can come with not having had children in hindsight. Can't say I have any answers for it. But it's obviously not as simple as this answer makes it.
If we don't solve alignment and EY is right about what happens in a fast takeoff world, it doesn't really matter if you have kids or not.
This IMO misses the obvious fact that you spend your life with a lot more anguish if you think that not just you, but your kid is going to die too. I don't have a kid but everyone who does seems to describe a feeling of protectiveness that transcends any standard "I really care about this person" one you could experience with just about anyone else.
Obviously no modern models are an existential risk, the problem is the trajectory. Does the current way of handling the situation extrapolate properly to even just AGI, something that is an open goal for many of these companies? I'd say not, or at least, I very much doubt it. As in, if you're not doing that kind of work inside a triple-airgapped and firewalled desert island and planning for layers upon layers of safety testing before even considering releasing the resulting product as a commercial tool, you're doing it wrong - and that's just for technical safety. I still haven't seen a serious proposal of how do you make human labor entirely unnecessary and maintain a semblance of economic order instead of collapsing every social and political structure at once.
Just because this isn't cheap relative to the world GDP doesn't mean it's enough. If our goal was "build a Dyson sphere" even throwing our whole productivity towards it would be cheap. I'm not saying there aren't any concerns, but the money is still mostly going to capabilites and safety, while a concern, still needs to be compromised also with commercial needs and race dynamics - albeit mercifully dampened. Honestly with LeCun's position we're just lucky that Meta isn't that good at AI, or they alone would set the pace of the race for everyone else.
I honestly think this is still cheap. Non-cheap would be monumentally bigger and with much larger teams employed on alignment to attack it from all angles. I think we're seeing Cheap and Fast, with the obvious implied problem.
I have seen this argument and am deeply sceptical this can happen, for the same reason why mobile F2P games rarely turn into relaxing and educational self contained experiences. Incentives are all aligned towards making AI girlfriends into digital crack and hooking up "whales" to bleed dry.
because an uncontrollable AI is not nearly as useful to companies as LW thinks
No one thinks it is. An uncontrollably warming up world isn't very useful to fossil fuel companies either, but fossil fuel companies can't stop to care about that, because it's too long term and they have to optimize profits on a much shorter time horizon. The argument isn't "evil companies profit from unleashing unaligned AI", it's "dumb badly coordinated companies unleash unaligned AI while trying to build aligned AI while also cutting costs and racing with each other". Cheap, Fast, and Doesn't Kill Everyone: choose only two.
My fundamental crux here is that this is ultimately going to result in more high-quality alignment work being done than LW will do, and the incentives for capabilities also result in incentives for AI safety and control
I don't know for sure if the infohazard concept is that useful. I think it could be only given certain assumptions. If you discovered a concept that advances alignment significantly and can't be used much for capabilities you should definitely scream it to everyone listening, and to many who aren't. But "this would lead to better alignment research" isn't very useful if it leads to proportionally even stronger capabilities. The goal here isn't just maximizing alignment knowledge, it's closing a gap. Relative speed matters, not just absolute. That said, you may be right - but we're already discussing something far different from "Great Man Theory". This is just a highly peculiar situation. It would be extremely convenient if some genius appeared who can solve alignment overnight on their own just starting from existing knowledge. It's not very likely, if the history of science is anything to go by, because alignment is probably harder than, say, the theory of relativity. But however unlikely it is, it might be a better hope than relying on collaborative projects that doom alignment to just keep lagging behind even if it in general advances faster.
because infohazards tend to connote the idea that there are ridiculously impactful technologies that can be found by small groups
I don't quite follow. Infohazards mean "some information is dangerous". This doesn't require small groups, in fact "it's safer if this information is only held by a small group rather than spread to the world at large" is inherently more true if you assume that Great Men Theory is false, because regardless of trust in the small group, the small group will just be less able to turn the information into dangerous technology than the collective intellect of humanity at large would be.
But much like with synthetic meats or plant based substitutes, this would require changing a whole tradition of brewing and distilling, with all the related culture and know how, with chemical synthesis. I don't think this is an impossible sell, but it would only stick if the benefits seemed significant. Butanol sounds like a very marginal improvement; people who dislike alcohol and its dangers will see it as no better, and people who like alcohol will see it as a poor substitute. Not sure if there's enough pull to actually justify the attempt to switch.
I assume they mostly evaporate when the paint is fresh, a time when it's pointedly not advised that you stand in the rooms taking deep breaths.
I don't think the concept of infohazard as applied to AI alignment/safety has anything to do with the Great Man Theory. If we bought the Great Man Theory, we would also have to believe that at any time a random genius could develop ASI using only their laptop and unleash it onto the world, in which case, any hope of control is moot. Most people who support AI governance don't believe things are quite that extreme, and think that strategies ranging from "controlling compute" to "making it socially disreputable to work on AI capabilities" may effectively delay the development of AGI by significantly hurting the big collaborative projects.
On the flip side, would alignment work proceed much faster and more successfully if thousands of researchers with billions of dollars of funding worked on it exclusively, building on each others' intuition? Of course it would. But we do not live in that world. We live in a world in which economic incentives have aligned things so that the power of numbers lies on the side of capabilities. In this world, if you share your genius interpretability insight (which you may have, because while science isn't made only of Great Men on average, it certainly can advance in random leaps and bounds depending on individual contributions, and here even a mere 10 or 20 years of difference in the time of a discovery can be crucial), well, it's much more likely that it will just be used to make an even more powerful AI before anyone manages to use it to align the ones we have. So, keeping info contained isn't the best thing to do in some absolute sense, but it may be the best thing we can do now, as inefficient as it is, because it's the only random edge that a small community competing with a much larger and more powerful one can hope to gain.
I will grant you that it is indeed possible that we don't understand enough about the brain to be confident that we won't just irreversibly ruin test subjects' brains with such a therapy, but this is much less likely than the possibility that either nothing will happen or that gains will be had, provided such a therapy is successfully developed in a way that makes it feasible.
The bit about the personality was specifically in response to the idea that you could revert brains to childhood-like plasticity. That's like an additional layer of complexity and unlike gene therapy we don't know how to begin doing that, so if you ask me, I don't think it would actually be a thing anyway in the near future. My guess is: most of your intelligence, even the genetic component, is probably determined by development during the phase of highest plasticity. So if you change the genes later you'll either get no effect or marginal ones compared to what would happen if you changed them in embryos - that is, if it doesn't also cause other weird side effects.
Experiments are possible but I doubt they'd be risk-free, or honestly, even approved by an ethical committee at all, as things are now. It's a high risk for a goal that would probably be deemed in itself ethically questionable. And the study surviving for example a cohort "gone bad" would be really hard in terms of public support and funding.
This reply is hilarious in the context of your first one. At first you confidently assert that changing genes in the brain won't do anything to an adult, followed by your statement that "we understand still far too little of how brains work" to know what's going to happen following such a therapy along with other predictions like total memory erasure. Which is it?
I mean, you sound like you know far more than me on it so I won't argue the specifics, but in general, "we know enough about this thing to not be able to safely mess with it, but to be reasonably sure that messing with it will have bad effects" is absolutely possible. It's in fact the default for really complex black boxes: while understanding their general function may be easy enough, if you don't know what you're doing messing with their internals, odds are that you'll break them rather than improve them.
The prediction of "total memory erasure" was a response to a specific idea, the notion that if intelligence was really mostly determined in childhood/adolescence, then if you could push the brain to regain its original plasticity you could repeat the process with different outcomes, and as I said that sounds like it would change a lot about what a person is (unless you can somehow disentangle experiences and personality from intelligence gained through them). I don't expect that to be the case if one of the premises doesn't hold; I was just criticizing that specific strategy. Others would not have this downside.
As for the rest, sure, it's possible that there might be gene changes that will simply improve neuron health. But if I had to bet I'd imagine it would be easier to gain traits like resistance to dementia and Alzheimer's or such by tweaking those, than a whole 40 points of IQ or such. I know brains aren't the same as ANNs, but to make an analogy, if you run GPT-4 on newer hardware it'll do the same things a bit better and faster, but it won't be able to make entirely new things out of whole cloth.
The original point was against using energy consumption as a measure of worthiness. It's true that all worthy things tend to consume energy, but energy consumption isn't proportional to worthiness, and some things that consume energy aren't worth anything at all. This holds whether one adopts a purely hedonistic view of utility or not.
But the refusal of wireheading is itself in service of a terminal value - because "satisfaction" is more than simple pleasure.
Not a good data point, they also have religious reasons for it. It also doesn't tell us which technologies are shifting the balance, if any (industrialisation? Addictive entertainment? Condoms?).
I think the strongest pro-Traditionalist argument is if those human connections were on expectation still more fulfilling, and they just happen to have a higher barrier to entry. I think that's probably fair (after all, you can always just play games and have watch parties with children and friends, and it's all the more fun!).
But even so, and even as I realise it, personally I still am married but childless and almost friendless. Why? A bunch of things, but mostly I would say a persistent lack of stability that has prevented me from establishing solid relationships and has made it hard to form new ones (the older you are, the harder it gets). It probably doesn't help that I am fundamentally introverted - that even as I enjoy company, it also tires me out. I have to wonder whether these are traits on the rise. It's one thing to rationally choose to be a loner and another to have your brain wired in such ways that the cost of reaching out is fundamentally higher. If such traits are hereditary, then they're likely being selected against as we speak, but if they're produced by certain education trends (e.g. helicopter parenting, or superstimulus induced ADHD) then they won't just go away any soon. Also, that lack of stability has roots in the way our economy works too, the costs of specific goods like e.g. an extra bedroom in your house, etc.
Right. See also: Zachtronics puzzles.
To steelman it, I think the "wasteful" position makes also a kind of sense: being pointlessly frugal when you're swimming in abundance can be the source of unnecessary suffering. But turning waste into a metric unto itself is begging to get Goodhart'd into doing some very stupid stuff.
The former thing sounds like overclocking a CPU. The latter instead "erase chunks of someone's personality and memory and let them rewrite it, turning them into an essentially different person". I don't think many people would just volunteer for something like that. We understand still far too little of how brains work to think that tinkering with genes and just getting some kind of Flowers for Algernon-ish intelligence boost is the correct model of this. As it often happens, it's much easier to break something than to build it up, especially something as delicate and complex as a human brain. Right now this seems honestly to belong in the "mad science" bin to me.
I feel pretty frustrated at how rarely people actually bet or make quantitative predictions about existential risk from AI.
I think that might be a result of how the topic is, well, just really fucking grim. I think part of what allows discussion of it and thought about it for a lot of people (including myself) is a certain amount of detachment. "AI doomers" get often accused of being LARPers or not taking their own ideas seriously because they don't act like people who believe the world is ending in 10 years, but I'd flip that around - a person who believes the world is ending in 10 years probably acts absolutely insane, and so people to keep their maximum possible sanity establish a sort of barrier and discuss these things as they would a game or a really interesting scientific question. But actually placing a bet on it? Shorting your own future on the premise that you won't have a future? That breaks the barrier, and it becomes just really uncomfortable. I know I'd still rather live as if I was dead wrong no matter how confident I am in being theoretically right. I wonder in fact whether this feeling was shared by e.g. game theorists working on nuclear strategy.
So for the most part I'm really happy with it - I think it's got a great UI and a great feel. I haven't much used the Dialogues feature (not even reading them), but they don't interfere in any way with the rest of my experience.
One thing I think might need some tuning is the feature that limits the post rate based on the karma of your previous posts. I've once found myself rate-limited due to it, and the cause was simply that my last 20 comments had been in not particularly lively discussions where they ended up staying at the default 2/0 score. Now I suppose you could construe that as "evidently you haven't said anything that was contributing particularly to the discussion", but is that enough to justify rate limiting? If someone was outright spamming, surely they'd be down voted, so that's not the reason to do it. I'd say a pattern of consistent down voting is a better base for this. After that I found myself trying to "pick up" my score for a while by going to comment posts that were already highly popular to make sure my comment would be seen and upvoted enough to avoid this happening again, and that seems like something you don't particularly want to incentivize. It just reinforces posts on the basis of them being popular, not necessarily what one honestly considers most interesting.
Why would this work on adults? The brain develops most in childhood. If those genes' role is to alter the way synapses develop in the fastest growth phase, changing them when you're 30 won't do anything.
I think the summary of my position on this would be that when we talk about "extinction" we tend to imagine more a violent event. Even if what drove that extinction was a splinter of humanity (e.g. a branch of superhumans, either biologically or cybernetically enhanced), that would still be bad, not only because it implies a lot of death, but because it means the ones that are left are our murderers, and I have a hard time thinking of anyone who would murder me as a worthy successor. If instead humanity gradually all morphed into something else I suppose taxonomically that still counts as the extinction of the species Homo Sapiens, but it's obviously not bad in any particular sense. We already do modify our bodies plenty: we do organ transplants, cosmetic surgery, birth control implants, Lasik, hormonal replacement and sex reassignment. We spend our whole life in a very out-of-distribution context compared to our ancestral environment. In all senses other than the genetic one, Homo Sapiens may already have gone extinct long ago, and we are already something else.
By the way, I am actually working on a post looking exactly at Jeremy England's theory (and at what I would consider its overinterpretation on e/acc's part) from a physics perspective. So look forward to that!
But the reasonable for being careful about this stuff aren't unfounded. Not only if we don't get precisely the function of each gene this could cause side effects of arbitrary seriousness, but the child is then stuck with them for life and would potentially pass them to their descendants. Now perhaps we should consider gene therapy more than we are now but it's far from a "go to the lab and start deploying it now" affair anyway.
Well, sure, there may be a more general argument for FDA bureaucracy being too convoluted (though of course there are risks with it being too lax too - no surer way to see the market flooded with snake oil). But that's general and applies to many therapies, not just genetic ones. Same goes for research costs being the big slice of the pie.