This kills the AI singularity with superhumans

>Gene therapy is easier than gene editing.
Editing is literally what CRISPR does, brainlet.

This isn't science fiction. The editing is already happening. All that needs now is for more genes to be ID'd.

As for investors, have you seen the markup on wedding stuff and baby stuff? Now imagine what people would pay not to have a retarded baby.

I know that, I use gene editing as a synonym for CRISPR. How can you call me a brainlet? Improve your reading comprehension.

What I said is still right, gene editing is happening but very slowly and with poor results.

>Now imagine what people would pay not to have a retarded baby.
Do you know know how difficult that is?

Do you have any expertise in this subject besides surface level pop science

Well, it's not that more genes need to be ID'd- we've actually ID'd most protein encoding genes in the human genome. The problem is in the nitty-gritty, like figuring out where in untranslated portions of the genome you find transcription factors. We've come a long way in creating databases of TFs, but it's still up in the air as to which each of them do necessarily. There are so many different factors that are working on any given gene, even incidentally, that it's incredibly difficult to control for how much of a given protein is ultimately transcribed and translated.

So in terms of trying to create superhumans, we're quite far away. Not to mention, none of us will ever personally see the benefits- modifying all of the somatic cells in the body isn't possible, so it would be modifying germ cells to produce progeny that have the target modifications. Then it's 9 months til you know if it even worked, even more if it's some phenotype that'll occur further along in development.

What's most likely the target for this technology, at least in the near future, is the industrial production of biological biproducts. With the ability to transplant genes, and the research into trying to get the structural component right, comes the ability to take genes from any organism and then use some model for production to grow it. We can do that with simple biomolecules like Insulin now- instead of harvesting it from pigs like we used to, we actually insert the gene into E.Coli and then grow it by the vat. Imagine such kinds of things, but with more complex structures, like leather, meat, ivory, bone, fruit, or any other kind of biologically produced complex- Then, it would be possible to take any product that is made by biological systems and then grow it on racks in factories. Even this is pretty far away, but there's a certain progression to our understanding of transcription factors and gene location that makes it more a reality every year.

Just curious, do you know what's being done in Chinese labs regarding human gene editing? I'm asking because I think a year or so ago I recall hearing there was some talk on the Chinese trying to reduce birth defects with CRISPR, and it was a whole ethical debacle, but I didn't really look into it at the time.

To add on to this, as I forgot, it's not necessarily just regulating the amounts of what protein via study of TFs, but rather studying interactions and post-translational modifications on proteins. To put into perspective how difficult this is, there's an estimated 20,000 protein encoding genes in the human genome (a shockingly small number, in my opinion), with roughly ~40-50 different types of post-translational modifications that can apply to many portions of a single protein. To figure out how each one interacts with each other, taking into account possible and likely locales for post-translational modifications and the kinds of concentrations you'll see them in, as well as the idea of multi-protein complexes that will interact with each other, it becomes an incredibly daunting task.

The other perspective for this task is that in order to figure out protein interactions, a common tool is to use NMR to print out what are essential topographic maps of how nuclei are shielded in a magnetic field. Take a collection of these NMR readouts with various changes to a protein- add a post-translational modification to see how the protein is going to bend/where it will latch on, or perhaps test the shape that two proteins that you know have interaction will form. It takes a lab or a group sometimes around 3 months for a SINGLE interpretation. The proteins are so big that they create so many fine gradients of nuclear shielding that even some of the best NMR machines in the world have trouble discerning, so you have to go in and make absolutely sure of what you're looking at.

There's also complexities within the genome itself; enhancers, promoters, insulators and let's not forgot epigenetics.

But yeah proteomics is hard shit, I did my thesis within it.

yeah that's what I largely mean when I'm talking about TFs. Bringing up epigenetics is tough, because pop-sci fags jump on it with no knowledge of it whatsoever, so I usually just don't talk about it unless it's super relevant.

Have a proteomics final on Wednesday. Shitposting about proteomics on Veeky Forums counts as studying, right?

The Chinese are apparently going all out with CRISPR and keeping research largely under wraps with the exception of monthly notes of success because the last time they came forward there was that ethics debate you mentioned. Meanwhile here in the states we are twiddling our thumbs over ethics and arguing over who gets copyright ownership of the method.

Getting my degree in molecular bio now, have a little over a year left to go. The idea then is either to move onto a genomics doctorate program, or to go to law school- I think I'll want to end up in law anyways, because this whole policy and copyright issue regarding the biological sciences and a post-genome world is fucking nuts. There's a huge vacuum for competent, science-literate lawyers, so I'm hoping I'm backing the right pony here.

Now I'm a red meat eating american patriot, but when I see shit like the fucking debacle over copyrighting genes and slowing down scientific/technological advancements for profit's sake, I begin to wonder if capitalism was a goddamn mistake. Not so much as to learn chinese, because working or living there sounds like a living hell, but the whole thing is fucked.

>Do you know know how difficult that is?
Look, you can dismiss what I'm saying as "pop science" all you want. You're in a position of expertise that has got you so close to the trees, you can't see the forest.

You mused about a lack of investors. Now that you realize that was an idiotic thing to say given the stakes, you deflected to "do you know how difficult that is?"

The answer is: it depends. In some cases, it's as simple as aborting a fetus that has a sonogram with thick folds on its neck. In others, there is no prenatal fix at all because it comes from later trauma or contamination.

Anyway, you never answered whether this supposed far away technology is, in your opinion, closer or farther than human-level or humanity-disrupting AI.

Both of these things (i.e. (1) radical genetic editing, replacement tissues, germ line customizations, etc. (2) powerful AI) are roughly on the same time threshold. Neither are science fiction, but they are not in the medium term.

I supposed you have a Ph.D. in AI also?

>modifying all of the somatic cells in the body isn't possible, so it would be modifying germ cells to produce progeny that have the target modifications

That was the point, m9. This was about competing with or merging with AI.

Not surprising. The Chinese don't give a fuck about ethics.

>Meanwhile, gene editing is still a fantasy.
no it's not you fucking retard christ shut the fuck up
it's about as far progressed as the neural networks who need billion-dollar research grants to make a robot pick up a cup, if you had to compare them

keras neural memeworks are not "AI"

>Chinese research
>valuable
Is this a meme or what

Developing AI that is actually "conscious" is even harder as it hasn't been done yet.