The article claims tensorflow uses eigen. That's not entirely accurate, although the pre-packaged builds of tensorflow are probably configured to use it. Tensorflow can also be configured to use BLAS/LAPACK at build time and not eigen.
Maybe you can install homebrew and open source apps to make it more Linux like, but you'll still be stuck with Mac OS's shonky window and task management UI unfortunately.
Install SizeUp. I paid $10 10 years ago and have been using it ever since. Far better window management than any Linux distro I've used. (and better than windows but that's not saying much)
I like Divvy because it supports more than just halves and quarters — I use a 7×6 grid so my browser can be wider than my editor and terminal: https://mizage.com/divvy/
Pairs well with Stay to make windows automatically return to their assigned layout when plugging/unplugging my external display: https://cordlessdog.com/stay/
> but it's much closer to an "interesting lab demo" than an "obituary for human programmers."
Was it being sold as the latter? I thought it was strange that a fuss was being made over this tech demo, and that explains why.
Parsing regular languages into other regular languages is exactly what transformer based LLMs should be good at.
A while ago there was an AI system publisized that was trained to generate animations from the game doom that was sold as AI created computer games. But the output made no sense if you watched for more than a few seconds. Isn't this the same kind of scare tactics dressed up as innovation?
This is not an acceptable use of gene editing IMHO. Cholesterol can be managed by diet. High levels of Cholesterol are down to choices made, not some inherited disease that patients couldn't avoid from when they were born.
High cholesterol is well documented to be heritable. Perhaps more relevantly, even if they would work, lifestyle changes have a significant patient compliance problem, which significantly reduces their effectiveness.
There is a more reasonable brother argument to could make, which is that we have well tested and effective drugs available today for managing cholesterol. Any new treatment would need to clear the bar of being better than those (in at least some circumstances) to be put into wide use. This bar may cleared by the fact that existing treatments often have adverse side effects.
Further, the one time treatment aspect is actually a demerit in some ways, as one cannot stop the treatment if there is an adverse effect. This means that the safety profile would need to be much better than us typically required. And proven over a longer timeline.
Of course, this is all concerns about approval and widespread deployment. We are still in the early human trial phase, where much more risk is accepted (subject, of course, to ethical guidelines).
Genetic treatments have the scope to not only have unintended consequences, but unintended consequences that can last over generations of people. I am in favour of them for some things, but we need to tread very carefully with the technology.
Maybe such a mutation would be a suitable target for a gene editing treatment. I'm not aware of all the issues involved there. I think the linked article doesn't have enough detail to form a fair opinion with.
Suppose that we were able to develop a drug that would allow people to regrow lost limbs through gene editing.
Do you think that such a drug should not be given to anyone who lost a limb due in an accident that was partly due to their own poor judgement because it was "down to choices made"?
Also, is there something special about gene editing that means it should not be used for these situations? Or if you go outside when it's icy and fall and break a bone, should the hospital refuse to treat you since that was your own fault?
I can't see how genetic modification would be the correct way to treat such injuries.
Maybe substances that can trigger epigenetic effects would be more relevant to such things. I understand a Japanese team is working on a means of triggering tooth regrowth by means of an injection. I've got no problem with that. Or something like Skele-Gro from Harry Potter either.
Why is it better for it to be managed with diet vs this? Presumably if managing it via diet alone worked universally they wouldn't be doing research into drugs like this.
I wouldn't go that far. Other, less invasive treatments should still be available IMHO, but there should remain an element of personal accountability. Gene editing is a very powerful tool, and messing with complex systems in powerful ways that we don't fully understand could be a recipe for many troubles down the line. I think the use of gene editing should be very surgically applied to obviously detrimental mutations, not for some scatter gun like approach.
What if the body raising cholesterol levels serves some purpose we aren't yet aware of? I've heard there's some evidence that medication to reduce blood pressure has a potential link to the onset of Parkinson's disease. Maybe messing with blood pressure in that way without addressing underlying causes has been a mistake, and messing with cholesterol levels without addressing underlying causes could also be.
That is why we do long and expensive trials before approving any medication for use.
Having said that, we have we been medically lowering people's cholesterol levels for decades, and the evidence seems pretty clear at this point that it is a net health benefit to those for whom treatment is indicated.
It is not at all obvious that targeted gene editing would be more disruptive to the body compared to flooding the body with a drug that happens to interfere with the one part of the process that we found a drug to interfere with.
Particularly if we are editing the gene to match a form that is already present in much of the population.
Some issues could only become evident over a period of hundreds of years with gene editing. That's longer than any medical trial I'm aware of. And mistakes made would be difficult, if not impossible, to undo.
If medications can already do what's required for cholesterol issues, why wouldn't we continue to use them rather than making some change to affect a complex balance that could cause problems over very long timescales?
If we were to be editing a specific gene to match what the wider population has, then I'd be more ok with that.
If you think so, what sources would you recommend? According to Wikipedia on medical ethics, "These values include the respect for autonomy". Not expecting any level of self control doesn't show respect for autonomy IMHO.
Interesting. I'll look into that. The Hippocratic oath says that a physician should do no harm (ἐπὶ δηλήσει δὲ καὶ ἀδικίῃ εἴρξειν). It's a personal value judgement as to whether some intervention is providing medical care or causing harm. I consider reckless genetic modification to be causing harm.
In the corporate environments I've worked in it is often company policy that all commits to source code control should have messages that start with reference codes to the coresponding ticket in the issue tracker (often jira). This how I look up the whys and wherefores of code changes.
A lot of places just don't do this. Unless there is a threat of a written warning some devs will do the bare minimum. I currently work with a guy that sometimes doesn't even run the code he checks in, I doubt he really knows how to code and everything is basically AI, management won't do anything and it the only remote job I could get. Another guy I work with won't do basic CSS fixes to things that are broken, so I cleaned up the login page for the site he is responsible for.
By whose judgement is the css you speak of 'broken'? Just making random code changes without a corresponding ticket is a recipe for troubles down the line. If you must do unprioritised work at the bare minimum create a issue tracker ticket to provide more information than can fit in a commit message. Something like selenium is good to automate acceptance testing (with screenshotting) cross browser as well.
> By whose judgement is the css you speak of 'broken'? Just making random code changes without a corresponding ticket is a recipe for troubles down the line. I
Obvious display issues in a mobile app on the login which is the first thing you see. Like the login not being on the page, things being misaligned (by quite a lot). Fixed in like literally 2 minutes.
This is basically due to laziness.
> If you must do unprioritised work at the bare minimum create a issue tracker ticket to provide more information than can fit in a commit message.
I create tickets for each commit I do. So people know why I have done things. These people don't do anything of the sort.
> Something like selenium is good to automate acceptance testing (with screenshotting) cross browser as well.
I can't even get these guys to write basic unit tests for React components.
There are two different worlds. There is the world where things are done more or less properly and there is the rest of us. I've heard of some places that don't even use source control, CI builds or anything that may have been novel 25 years ago.
Has anyone considered decompiling eloquence? With something like ghidra or ida pro? Mario 64 was turned back into high level language source code this way.
This wouldn't be easy due to Eloquence's internal architecture. eci.[dll|so|dylib] only contains the low-level platform abstraction layer, things like threads, queues, mutexes etc, as well as utility classes for .ini file handling and such. It then loads a language module (from a path specified in eci.ini). The actual speech stack is statically linked separately into each language module (possibly with modifications, not sure about that); in theory, if you reverse-engineered the API between the main and language libraries, you could write an Eloquence wrapper for any arbitrary speech synthesizer. This means you'd have to reverse-engineer this separately for each language.
From what we know, Eloquence was compiled in two stages, stage1 compiled a proprietary language called Delta (for text-to-phoneme rules) to C++, which was then compiled to machine code. A lot of the existing code is likely autogenerated from a much more compact representation, probably via finite state transducers or some such.
I'm bullish on LLMs being able to help with this kind of reverse engineering effort, if not current models then in a few more years. I've had conversations with people where they managed to get Claude to help reverse engineer old weird binaries with very little input. I wouldn't hype it up as being a magical tool that'll definitely work, but it can't hurt to try.
I gather decompiling mario 64 wasn't easy either. Just having C++ that can be recompiled to other architectures would seem to be useful. The original Eliza chatbot was converted to modern C++ in a similar way recently, and that used a compact representation for its logic as well.
My experiences with ONNX have not been pleasant. Conversions from models written with Tensorflow and Pytorch often fail. I recommend using TFLite or Executorch for deployment to edge devices instead.
Agreed, I have seen some speedups with ONNX if I'm being honest but the process especially on MacOS is a bit messy. I'll try out Executorch and see how it compares, cheers for the recommendation
