qsera

I do this, or in other words, I had to do this to go on. I think everyone comes up with some philosophy to keep on going. Even a street bum might have come up with philosophy on par with some of the greats that keeps them going...

In fact, for this particular idea, I am so over the edge that I sometimes feel pity for those conventionally lucky people who seemingly have it all and don't have a struggle on their on. I mean, what is the point of their life? But I find consolation in the possibility that they might have hidden struggles...

Was hoping to contain more in-depth content...

There is no form of evidence generation, real or hypothetical, greater than the randomized controlled trial.

Wrong. You are lying by omission.

Randomization and controlling are essential. But the version of RCT that we currently employ have problems with control and evaluation. If you have a flaw in control or evaluation, then the whole thing becomes flawed.

So the idea of RCT is good, but we are limited by our ability to control and evaluate.So there can be a better version of RCTs that we currently employ, but with precise control and evaluation.

but the beauty of statistics is we actually can know

The beauty of statistics is also that it is quite easy to hide falsehoods behind statistics.

There Are Three Kinds of Lies: Lies, Damned Lies, and Statistics

(I am not even going into the fraud problem, which is what we have discussed before, which turned out to be quite useless)

I am not sure, but the claim was that there is some date based correlation when this drug started being used, and rise of some health issues kids...

Why would it need to?

Because that is what everyone (I mean, not if you are pharma) is interested in...

highest quality evidence generation system we could possibly have, which is RCTs.

The "best we have" does not mean it is good enough for the task at hand. Just saying.

Basically what I am saying is that qualifying something by saying "best we have", does not justify its using on its own..

clean chit!

They can spot contradictions in the the prompt. But not in their own training data.

biased where AI does not.

AI can be totally biased...

The fact that it can spout bullshit all day long to a human who can be tired and would actually act on the said bullshit, is not very comforting...

For example, an LLM could confidently declare something a tired human would take as a fact, but would backfire in a real world.

what stops an LLM from applying similar lines of thought* in a manner that results in a novel conjecture?

One thing is that an LLM can never assume, or find out, an inconsistency in its training data. Novel ideas often require correction of existing assumptions. As far as I understand, it is impossible, by design, for LLMs to contradict what is in its training data.

For example, an LLM trained on the data from an internet comprised of people who believe in the earth centric hypothesis can never say "Hey, that cannot be correct", or come up with the heliocentric alternative

But maybe it is not applicable to pure Math...

From the post lol

So I wouldn't really say that this result is using or creating some fundamentally new techniques in convex geometry or optimization theory. What this means from my perspective is that if a result is attainable with existing techniques, modern AI methods will be able to solve those problems. I don't think researchers in math/TCS will be made obsolete, but I think it will instead no longer make sense to work on any low-hanging, or even medium-hanging (you know what I mean) fruit. We'll be needed for problems where actual novel approaches are needed.

because of AI, I've completely lost interest in this kind of stuff.

Where did the joy for doing it came from, before LLMs?

I mean, did people stop playing with LEGOs because we now have 3D printers?

Either you didn't really "enjoy" making stuff before, and were only doing it as a means to an end. Or you are just misguided for a bit, and will come back to it a while later..

It literally takes the point away since folks will simply say "oh, who cares? Look at this shit I made in AI instead"

But that does not make any sense what so ever. Who cares how you did it.

I think people are going to me more sympathetic if you said that you didn't use LLMs to build it.

I've been coding since I was 11 and loved every minute of it

Can you share some of the stuff you have made (before LLMs)?

Those were all potential futures we imo thankfully and carefully avoided!

I am not so thankful. But it is coming from my view of the current world.

I think open Internet for business is good. But open Internet for communication was bad. I cannot think of one good thing that have resulted from an open internet for communication/discussions. It has turned out to result in a collective, perpetual state of seizure for all humanity.

I am someone who had to learn stuff from random encounters with books in the 90s. Even then, I would not have it any other way...

I think it is better to have some elitist communities, than to drag everyone down to some least common denomination...

I think that was one of the big mistakes/oversight that came with the Internet..

Honestly, horrible example.

Can we really be sure that thoughts, opinions, taste will still come out sharper and faster after five, ten, 20 years of using these tools almost every day?

After 5 years, I think the thought profile every power user of the LLMs would be an LLM derived carbon copy of each other.

Prepare the world to get even more boringly uniform

The concrete issue is this. We seem to want to give every incentives to these companies as possible to get them to drive the EV transition as soon as possible. Even when these incentives are in the form of free pass in terms of lax regulations that I mentioned earlier.

You are asking me to give a concrete example of how it will be abused. I don't have that. Sorry. You win, I guess.

There could be some unknown unknown which makes these actually bad..

No, it is not the unknown unknowns that worry me. It is a very much known aspect. It is "human behavior" and the collective short shortsightedness that goes along with it..

Combine that with any potential issues that EV tech have, and we would find ourselves in a very similar or far worse mess than we are now.

So I want us to be collectively better than that. But the responses that I am getting when I raise this point, does not give me much hope. Because people latch on the easiest thing. That I am "Fear mongering"!

On concrete terms, this would look like some regulation that require EV companies to declare a plan regarding how they will process/recycle old batteries before they launch a product. And a way to closely watch that it is being carried out.

I don't see any efforts or calls to do things like that. As of now, anything "green" gets a free pass.

you should be worried about the fumes released every time you start up a fossil fuel engine...

I am worried about both. Why are you picking sides? At lease fossil fuels started slow. Here we are switching to an unknown solution(s) on a massive scale..

but they are batteries with 10 to 20 years of service life as opposed to ~1 week worth of fuel transported, refined, and burned. Yes they pollute, but they pollute a lot less, particularly over their lifetime.

CO2 at least have a natural mechanism by which it gets removed. And the affects of CO2 accumulation is not as acute. I am not sure about the materials that potentially could be used.

Also you say they pollute less. But as I said, the devil is in the details. Remember that the lead that used to get added to gas turned out to be very bad. So may be it is something like that that would be added to the battery chemistry to make it last longer or work better in cold or what not...

It is disturbing that people are just handwaving such possibilities away. I think that was exactly how we got here. But now doing the same could end up a lot more worse...

But scale matters too. I wouldn't be too worried if we are not going to eventually replace all fossil fuels with batteries.

That is a lot of batteries..

LLMs be able to coordinate their understanding

I think the next time I see "LLMs" and "Understanding" in the same sentence, I am going to lose it....

You paint a happy path. But the devil is in the details that often come to light only much later.

Sure, we could end up with a tech that have really no pollution potential. But there is much higher chance that we would end-up with something bad, that would go into production with a promise of "proper handling and re-cycling", but would never end up getting done...

One that makes sense.

Please allow me an opportunity to enlighten you:

So I am aware. The "we" is important because I said "accurately". When "we" set up the recording, we can make it to record with the required precision. Inferring measurements from incidental things can give the required precision, but it should be justified.

Polluted cities getting cleaner for some. Clean rivers and beautiful forests getting poisounous ore pits for others.

Yes, that is how it goes.

Right now all the "intellectuals" and "thought leaders" are in EV apologetic mode. This is a window of opportunity for the corporate to maximize bad stuff to maximize profits selling EV.

without good reason

That applies to things you're experiencing on a daily basis.

That we cannot accurately look into the past unless we explicitly record it, is something that I experience on a daily basis.

Seriously, if you are so convinced, then why don't you just give me an accuracy measure and reasons to support it? I am not interested in anymore articles that does not contain an explicit accuracy claims and reasons to support the same.

So please understand this. It is not that I am questioning the existence of such techniques. I am questioning their accuracy/reliability..

"how much they last" does not play a part in the recycle load the industry will have to handle though...