Hacker News Reader

Comment by jpgvm | Hacker News Reader

Man this comments section is a cesspool.

The reality is that consensus does play a part in science, this is because intuition also plays a large part - both in guiding experimentation but also in interpreting findings.

Consensus is the mechanism used to translate scientific findings into actionable advice/recommendations/engineering. Without it nothing would ever happen. Science is always full of unknowns, consensus allows us to move forward in absence of complete information (because such a thing often doesn't exist).

The means by which scientists and engineers reach consensus are very different from politicians and other soft studies. Because of the role peer review plays in the scientific community consensus is built up over time, not through discussion but rather through experimental results and verified data. Consensus is less "we sat around and decided X is true" but rather "there is an overwhelming amount of data that suggests this is true". This is often translated into models which spur further experiments to attempt to verify or disprove these models (for instance the LHC and the Higgs boson as a high profile example).

So no. 99% consensus in science isn't that uncommon and no it's generally not a bad thing. A big difference is also that in science it only takes a single very compelling counter-point to significantly disrupt consensus. See here Einstein's photo electric effect or his theory of relativity - both disrupted fairly intrenched consensus.

This latter point is very important because unlike scientists soft fields often struggle to break consensus even in the face of overwhelming evidence. Everyone feels like they should get an "opinion" and that scientists are just voicing "opinions". This leads to people equating the voice of relatively uninformed individuals with that of scientists. I don't know this can be resolved but if it could be it would allow society to move forward at a much faster rate.

View on HN

We have to acknowledge the fact that the denialism concerning climate change extends from it being strongly linked politically. That's the reason we see so much effort in this thread concerning "valuing other opinions" or "people believed the earth was flat". Since the naysayers aren't interested in coming up with legitimate alternative models, the only refuge is to downplay "consensus" and fight to have uninformed individuals on the same level of discourse as the climatology community. So while the science is solid, you aren't going to get someone to believe something they are against from the onset.

> Since the naysayers aren't interested in coming up with legitimate alternative models, the only refuge is to downplay "consensus"

It's important to note that this pattern plays out for many issues. It's reactionary politics and rhetoric, which is simply against things and for nothing - to deny: climate change, covid, racism, etc.

It's important to note that this pattern plays out for many issues. It's progressive politics and rhetoric, which is simply for any change and refuses to admit that anything is ever good enough, has natural causes, or is a completed process - to blindly accept claims of: climate change, covid, racism, etc.

This cute comment would have more bite if you'd actually managed to write in a straight logical line like the parent comment. Of course, imitation-but-worse is a reactionary staple.

This is probably my favorite response, because I literally made the exact point as the original comment. Any difference in the quality of the argument, such as it is, is purely invention on your part.

Which of course, is the point: it's a stupid argument however you look at it. If you up vote it, it's only because it confirms your priors, not because of the quality of the point (non-existent).

Strange that both the coal and oil industries (independently) produced internal documents in the mid 1960's agreeing that climate change science was sound and thus it was a particularly serious threat to their business.

Are you saying that they have not supported climate change denial?

“Non-movement is the same and as valid as movement!”

Sure, good luck with that, but why impose it on others?

This is a well known PR tactic, described in Merchants of Doubt and most recently in The New Climate War by Michael Mann.

Interesting enough, this tactic can also be deployed on strong believers in climate change in several ways. For example, the very idea of a carbon footprint is invented by BP in order to shift responsibility towards the consumer, which deflates pressure on corporate action and regulation. The gun lobby, tobacco companies and softdrink industries have employed similar campaigns - such as inventing recycling.

The very term 'climate change' is a friendlier and successful rebranding of global warming by the way. It suggests something natural, maybe even benign. That's why I prefer climate crisis.

Another example is the idea that renewables are not 'green', like wind turbines kill birds, EV requires lithium etc. The proposed alternative is always either some future technology or a complete non-starter like dismantling capitalism and just stop using electricity.

Yet another tactic is to make the case that we are doomed and no action can be taken.

All these serve the interests of the fossil barons very well as they divide, cast doubt on, delay or otherwise frustrate the efforts to genuinely decarbonize our economy.

Thus, one needs to be careful to distinguish between actual science and marketing disguised as science, or more insidious: bits of science employed in a psyop campaign.

I would argue that the recyling example is one that they're still winning the PR war on, very similar to the carbon footprint one.

Recycling is green and economic. This is the scientific consensus and also blatantly logical. Anyone arguing otherwise has consistently been in the pocket of big business or political parties that thrive on letting big business do whatever they want. We have decades of evidence of that.

But the most common story you hear about recyling is that it's a pointless greenwash and the alternative is ..... the status quo. Constant stories of how it went wrong in some case, or got subverted by those same big companies and the moral always seems to be "don't try to do green stuff, it's all a scam and/or silly hippies who do more damage than good", which is exactly the "doomed" strategy you mention.

Almost certainly a whole bunch of people reading this comment have a direct mental connection of "recycling" -> "scam", "waste", "not even good for the planet" similar to what has been cultivated for wind turbines, solar, EVs etc though most of those seem to be losing steam faster, probably only because those industries have their own lobby groups now.

My comment about recycling needs more nuance. Of course recycling as a mechanism is part of a sustainable economy. But much better to recycle a material is to reuse a product, and even better is for the thing to not have been produced in the first place.

Recycling plastic for example is very limited. It's very hard to get everybody to perfectly recycle, then we can still only recycle a limited amount of plastic a limited number of times. Clearly, recycling plastic isn't green. It's only ever so slightly less harmful.

We have been recycling plastic for a long time and we are still choking in it's waste.

I still recycle myself though, even as I'm trying to reduce my consumption of (single-use) plastics, which isn't easy at all. But I also support a ban on as much plastic products as possible, since recycling is a non-solution and the problem unacceptable. I think such bans are gaining public support, at least in Europe. Just like the public is starting to become aware that individual action on reducing footprint isn't enough and even distracts from holding the right actors accountable. Landmark cases here in the Netherlands such as the Urgenda case and the case against Shell help spread that awareness.

I urge you to examine where you got the idea that recycling plastic is such a bad idea.

"Biomass is so bad it's worse than coal" "EVs are so bad you're better of with an ICE" "Recycling is so bad you're better off not bothering"

Notice a pattern there? Attacking alternatives that are better, but not perfect (nothing is) is an obvious tactic to fall back on once the more blatant denial starts to falter.

Meanwhile, if you looked up a study about recycling, including specifically plastic, you'd see the consensus is that it saves energy, and carbon and money.

There's definitely nuance but plastic is unavoidable in modern society. No one is going to ban it entirely, even if they do ban most disposable uses. And if they don't ban it, then the people who create products with it should be forced to pay for the recycling and someone should check it actually happens. There is no downside to this. The extra cost being borne by the producer even makes it less attractive to people who only care about money.

Ignoring that easily achievable collective goal in favour of an impractical personal achievement of avoiding plastic is exactly the kind of thing people seem to be waking up to these days.

Reduce reuse and recycle. Yes they all matter, but not bothering to recycle the stuff you don't reduce or reuse is obviously only beneficial to the people who sell the fossil based input materials.

You are pulling a straw man I've even been careful to avoid since I noticed it:

> you're better off not bothering"

Nowhere in my comments did I say that it's better to not recycle. I've even made the same argument as you did regarding EV elsewhere, debunking the idea that because EV isn't perfect, it's not worth the effort.

I've said "I still recycle myself" and "Of course recycling as a mechanism is part of a sustainable economy". I do not have, expressed or implied, anything like that recycling plastic is a bad idea. I don't know how to make that point clearer. My point however, is that it isn't a solution and convincing the public that it is, was a deceptive corporate propaganda campaign by the leading beverage and packaging corporations, designed to avoid taking responsibility and more drastic actions.[1] It's painful, but the green movement must recognize it has been misled and that recycling is really not enough, not by a long shot.

> plastic is unavoidable in modern society

This doesn't mean we can't significantly reduce our consumption of single use plastic and avoid a substantial amount of plastic usage, which is only increasing.

[1] https://www.chicagotribune.com/opinion/commentary/ct-perspec...

I know you understood this for EVs and renewables, that's why I pointed out that I think you have fallen for it when it comes to recyling plastic.

You said this:

> Clearly, recycling plastic isn't green. It's only ever so slightly less harmful.

You didn't need to emphasise how little you think recycling helps. I take it to mean "it's better to just not use plastic" and agree but as I said, on a society wide scale that is simply not happening so advertising that attitude to other people, who certainly will not stop using plastic, spreads the meme that recycling is basically useless and that is a bad strategy. Which coincidentally is the same message that fossil fuel providers push.

Collectively we need to use less and recycle the stuff we do use efficiently. As I said, there's no downsides to this, nor any conflict between those two things any more than saying we should use both wind and solar.

You don't use solar when wind makes more sense nor vice versa. But if you personally decide to install solar panels because that works well for you, it's not helpful to say "wind is not green, just ever so slightly less harmful". We need both, both are better than the status quo.

The difficulty is that recycling is a mixed bag. Recycling of aluminum is a resounding success environmentally and economically. Steel recycling seems to be a pretty good deal for everyone, and last time I checked glass was environmentally good and economically break even.

Unfortunately the term and concept of recycling covers all of these plus plastics, which have been largely a failure. This confuses both consumers and policy makers.

This is a "motte and bailey" rhetorical trick they use.

The people against any kind of market regulation will argue that all recycling is inherently inefficient if the market doesn't do it spontaneously.

This is obviously undermined by the obvious success of many metal recycling schemes where the regulations created a market.

So they then flip from recycling to "plastic recycling". If you then point out the many plastic recyling successes they'll jump to specific plastics and ways of using plastics that make it harder to recycle them. The obvious solution of using further regulation to get people to update their processes to use the easily recyled plastics is mysteriously avoided even though it has literally been done for decades across the globe, sometimes led by corporate co-operation (e.g. milk in various countries is in semi-standardized containers).

If you have a gut reaction that recycling, or just plastic recycling is a bad thing, you've been conned by fossil fuel producers who compete with recycled plastic as a feedstock.

If anyone is interested in the details of the history and mechanics of fossil fuel propaganda, this [0] podcast is a good resource.

[0] https://podcasts.apple.com/us/podcast/drilled/id1439735906?i...

> So no. 99% consensus in science isn't that uncommon and no it's generally not a bad thing.

To put things in perspective, there is a 97% consensus on evolution theory: https://en.wikipedia.org/wiki/Level_of_support_for_evolution....

That basically means we are more certain about the cause of climate change than about evolution.

> Consensus is the mechanism used to translate scientific findings into actionable advice/recommendations/engineering. Without it nothing would ever happen. Science is always full of unknowns, consensus allows us to move forward in absence of complete information (because such a thing often doesn't exist)

FWIW, my experiences have been the precise opposite. My colleagues never seem to have consensus about anything. But companies move forward with our recommendations, both because we're more likely to be right, and if our recommendations happen to serve the company's goals.

I'm sure a bunch of people will read into this comment, but I am speaking purely about my experiences as an ML scientist. It would be valid to say that ML represents a very small fraction of the global scientific apparatus, and I certainly don't claim to know what it's like to work at CERN.

It surprised me, though, how often you were right -- there often is consensus, in the sense that everyone reads about a certain technique in papers and automatically assumes it's applicable to their own situation. For example, everyone in language modeling almost universally agrees that "lower loss is better," ("loss" is a specific ML term,"). But that's only true for general language models trained on large quantities of data. If you're fine-tuning a language model, you almost never want the steepest loss slope, because it means your learning rate is too high and you'll destroy the existing knowledge.

I would say that experience matters, and each situation is different. It's super tricky to make an argument that consensus is both universally common and universally good.

ML has gone through a significant enough transformation in the last few years that it's velocity essentially makes it very difficult to form consensus about anything right now as everything feels up for replacement in short order.

That's probably true, which is why I love it so much. :) You're right that it might not be the best example.

Still... in particle physics, for example, I've heard that there was consensus in favor of certain particles existing, even to the extent of "multiple experiments have achieved an N-sigma result showing that this particle exists," for some fairly high N. And then a later experiment showed that the particle simply didn't exist, and that the previous results were experimental error. (Specifically, experimental bias – they were expecting to see certain results, and so they were probably subconsciously combing the data for such results.)

My knowledge on that subject comes from Veritasium, not from reading physics papers or following along with experimental breakthroughs. But I've seen similar things happen in ML, where people are pretty convinced a certain technique is good (hello, style mixing in StyleGAN) and no one ever bothers to check whether it's true in reality (style mixing is what prevents StyleGAN from being as cohesive as BigGAN).

I guess I broadly agree with your point. There is consensus that black holes exist, and that we've even observed one, so it's pretty likely that if you were to travel there, you'd find a black hole. But given how much science itself has changed throughout history, alarm bells start going off when hearing things like "Consensus is extremely common, and it's almost universally a good thing."

"Consensus is extremely common, and it's almost universally a good thing." isn't about saying things are assumed when they are not, just that the community is able to quickly adapt to changes in fundamental theory in order to explore the new ground opened up as a result.

Take your example about StyleGAN vs BigGan, I assume once it became clear that the latter was superior to the former that likely resulted in changes to existing architectures that then found additional improvements. This change in consensus is what enables that and is a good thing.

I think the problem here is conflating consensus with faith, that is assuming truth in absence of data. Consensus is easily changed with the introduction of new data, faith hangs on no matter how much evidence is put forward that it's horseshit.

> Take your example about StyleGAN vs BigGan, I assume once it became clear that the latter was superior to the former that likely resulted in changes to existing architectures that then found additional improvements. This change in consensus is what enables that and is a good thing.

Well, no. :) But I think the "no" is because of the uniqueness of ML rather than a "no" to your point in general. You might be right about other fields; I don't have experience there.

In ML, there an enormous number of techniques. Style mixing was presented as a core feature of StyleGAN (https://arxiv.org/abs/1812.04948) and was enabled by default in the codebase (https://github.com/NVlabs/stylegan/blob/03563d18a0cf8d67d897...).

So there's a lot of "inertia" -- for example, when StyleGAN 2 came out, style mixing was still the default (https://github.com/NVlabs/stylegan2/blob/f2f751cdc7f996e3138...).

I haven't had time to dig into StyleGAN 3, but I suspect that style mixing might still be enabled by default.

It wasn't until we did a detailed, methodical analysis side-by-side with BigGAN, specifically to answer the question "Why is BigGAN so much better for diverse datasets?" that, on a whim, I turned off style mixing and was astonished to see BigGAN type quality pop out of a StyleGAN type arch.

Discoveries like that usually go unnoticed, frankly because it's a lot of effort to write a paper specifically to say "Hey, if you're training StyleGAN, definitely turn off style mixing. It only seems to work well on faces."

However, if such a paper were to be written, and accepted into a peer-reviewed journal, then your original point would probably be valid. So I don't even know if it's worth writing all of this -- I just thought it'd be interesting to point out the "Well, not really" in this case. The knowledge ends up floating around on Twitter and Discord rather than being transmitted via scientific papers...

But, this all does tie in to your final point:

> Consensus is easily changed with the introduction of new data, faith hangs on no matter how much evidence is put forward that it's horseshit.

It's remarkably easy for old, accepted ideas to hang around. You'd think it'd just be a matter of "Run the experiment; experiment proves thing; thing becomes accepted." But in practice it's felt quite different...

The thing is, everything you're saying is true in general. As t approaches infinity, there tends to be more and more consensus about older ideas, like the existence of black holes, or the validity of laws like F=ma. So we should probably pay attention when there is 99% consensus on a particular topic.

But, for example, one reason I wouldn't want to publish a paper claiming style mixing was bad, is because it would contradict the results of Karras, who is famous. I'd better be very certain about my claim! So there's sometimes a reluctance to contradict the consensus, too, which ends up equivalent to "faith" in your example -- we have faith that famous scientists are correct. (They usually are.)

As a cherry on top, I'll just leave a link to Feynman's messenger lectures: https://www.youtube.com/watch?v=-kFOXP026eE&ab_channel=TalkR... ... the history of science is fascinating. I'd dreamt for years of becoming a scientist, but the actual experience turned out to be surprisingly different than what I thought it'd be. I love it though -- all these weird corner cases are the spice of life.

>But I think the "no" is because of the uniqueness of ML

In a thread where people are parsing the meaning of consensus with respect to climate change, citing the atypical example of the state of consensus in a unique field is going to contribute to the loss of context that makes people confused about the original subject at issue in this thread.

It's great to know about BigGAN vs StyleGAN, but I'm almost positive people will skim this detour into increasingly detailed elaboration on machine learning and think the upshot is supposed to be "and therefore we shouldn't put stock in climate science consensus."

Yeah I think everything you wrote here is on-point, nothing here happens instantaneously and with the current velocity of ML it's even less clear that it's taking place.

Thanks for your really in depth and informative reply though, I'm only cursorily interested in ML through its relation to data, which is my speciality. My exposure to ML is mostly through said ML teams making requests for data and systems like feature stores to run their models so I have some understanding of how they construct their overall architecture but not the specifics so this was super interesting.

I'm not sure what particle physics experiment you are talking about, generally it is very difficult (or even impossible) to show that something does not exist, so I think the story likely went somewhat different (although the gist of it might still be correct). I'm not a particle physicist either though.

Regarding your comment on ML and consensus, I think there is a bit of a misconception of what scientific consensus is. I would say it is much broader than the examples you give. ML is essentially terrible example to argue about consensus, I would say there isn't even consensus about what ML even is, let alone any of the details of how you train a network.

Regarding your earlier example about loss, I'm a bit confused. Isn't it that in optimisation theory the goal that you want to minimise your loss (so lower loss is always better)? Now that might not mean that you should travel along the steepest gradient (this might lead you to a local minimum) and your loss might not represent what you actually want to optimise for (i.e. your loss function is faulty). But from a fundamental theoretical point of view, you always want the lowest loss?

Unfortunately, it's 2:45am, and I have a meeting tomorrow. But I'll try to set aside some time for a detailed reply.

The cliff's notes are: the first time I was able to train GPT 1.5B on gwern's poetry, I was super proud of myself, because it was extremely hard to get 1.5B to train back in those days. So we let it train as long as possible, and the loss kept dropping lower and lower. So we'd periodically dump samples -- surely we'd get better poetry, right?

No. Eventually the model degenerated into broken-sounding English. Because our training data was 50MB, so it quickly forgot how to speak English at all -- the prior knowledge got wiped out, and was replaced by that 50MB dataset.

The term for this is "overfitting," and it's remarkably easy to run into it in practice. The cure is to have more data. But how do you gather more than 50MB of poetry? That's a lot of poetry. So what do you do when you want to train a poetry model?

The answer is to train a general model on a massive dataset (https://arxiv.org/abs/2101.00027), and then to fine-tune it on a smaller dataset (like 50MB of poetry). But during that fine tuning step, you'd better watch out, because everything you thought you knew about ML is wrong: You need to have a gentle downwards-sloping loss curve -- steep curve is terrible, since it means you're destroying the old knowledge without giving it a broad view of the new knowledge. And you need to stop before hitting loss < 2.1, otherwise you'll be overfitting and the model will suck.

I suspect that hundreds of great ML models were overlooked for these reasons. There are hundreds of thousands of people messing around with ML models for fun. Lots of them try to train those models on small datasets, because small datasets are all you have when you start out. So they google how to do it, hear "Low loss is good," train overnight, dump samples, and then end up going "Oh well, the model seems to suck, but the loss is low, so I guess I didn't do it properly." They probably did train it properly, but they weren't continuously dumping samples, so they didn't notice they needed to stop before the model turned into word salad. (Or art salad if it's a GAN.)

I would say that from a fundamental theoretical point of view, "you always want the lowest loss" is only true when you have an infinite amount of data in a target domain. But in practice, you probably have a few gigabytes if you're lucky. Even imagenet is only 1,000 different image classes, which means it's possible for a model to memorize those 1,000 classes and overfit, rather than achieving a general "understanding" of different kinds of objects.

I got all of these details wrong when I first started out, so it was really eye-opening to see that it's a terrible idea to rely on common knowledge in ML. Common knowledge is an excellent starting point, but you quickly learn to think for yourself, or you just end up imitating what the big players are doing while wondering why none of your models seem to be working.

(... I guess I wrote a detailed reply after all. Ah well, sleep is optional as long as you choose to ignore the effects of sleep deprivation. That's how that works.)

Ok, thanks for the clarification. I agree if you are overfitting lowest loss is not always best, and I did not think about that aspect when I wrote my reply.

I think with everything you describe is one of the main "gripes" I as a more traditional engineer/scientist have with ML. It often seems like a giant black box that I know is generally vastly overfitted and it is often extremely difficult/impossible why we get the solutions that we get, i.e. we don't really get scientific insight. Nevertheless I realise it is an extremely powerful tool.

Hope you can catch up with some sleep today, sleep deprivation effects can only be ignored for so long (especially as you get older as I have noticed).

I think you really need to name a couple particles like that if you're going to bring it up.

The most biased super-precise physics I can think of offhand is when someone measures a physical constant slightly wrong and then subsequent results all lean in the same direction. This happens more than once, most notably with fundamental electric charge, but it's absolutely nothing compared to a nonexistent value!

https://en.m.wikipedia.org/wiki/List_of_experimental_errors_... has https://en.m.wikipedia.org/wiki/Oops-Leon? But that is a single experiment which very specifically did not have a high sigma result.

Also the vast majority of companies fail.

I’m not sure what your point is, unless it’s saying that a lack of consensus is indicative of being wrong.

Which would support the commenter you replied to’s argument.

So many word for so little content. We are mostly scientists here so you don’t need to be so condescending. Problem is that your words are just opinion with no basis in reality, no consensus get to 99% without some manipulation.

I am very confident that HN posters are not "mostly scientists." This community is largely skeptical of even undergraduate degrees, let alone PhDs and academic publication. What percentage of posters do you think have published a scientific paper? 2% would strike me as high.

Consensus is basically what Peer Review is. Having a bunch of other scientists look at your methodology and data to see if they come to the same conclusion you did, or if there was a mistake you overlooked.

This 99% is basically like a meta-peer-review of all peer reviewed papers.

I think the larger issue is the lack of credibility. Having read many of the papers, been involved in some research myself, etc. I thoroughly believe the field has no credibility. We simply don't know and it's unknowable depending on what we are attempting to measure, predict, etc.

Here are the criticisms I have seen and share myself:

1. Define "climate change"

2. Funding tends to be provided to those who agree politically

3. Funding is typically tied to particular research objectives. Because of that, outcomes are often per-determined, particularly when we're looking at trend lines in particular predictable ways. For instance, we are likely funding research into "what are the human causes of climate change" -- which implies an outcome.

4. While there may be consensus that it's human caused, that doesn't necessarily mean there's consensus on the reasons, outcomes, scale, etc.

5. Journals / Reviewers actively block publication of counter points.

6. There's significant evidence climate change will benefit humanity, aka the characterization from politicians misrepresents the actual data. This reduces the credibility of the field and influences desired outcomes.

7. Many of the predictions from scientists in the field have been wildly off. If you go through old papers you can see there were predictions of an apocalypse every few years, which never comes. This makes it seem like a giant grift.

8. There's a MAJOR sampling issue. Where are the data samples taken to determine "warming", there are many papers cherry picking data. Certain regions are getting cooler, others warmer, etc. Additionally, we only have data reliably for a hundred years and we're exiting an ice age. We see correlations at that scale, causation are not discoverable; at least for what most people consider climate change (global temperature change & weather patterns).

9. Obvious counter examples exist. Per-industrialization, North Africa went fro a lush land to a desert. It wasn't human industrialization that did that. North America was covered in a sheet of ice a mile thick 20,000 years ago. These events are rarely considered in papers examining human influence.

On a quick glance I see numbers 1 ,43,44,48,90,96, of the climate change deniers list (1), and that's just from a 30 second skim. You are doing the classical playbook of "just asking questions" which either can't be settled (but what really is climate change?) or have been settled and you are willfully ignoring them (the IPCC has a clear definition of climate change, anthropogenic climate change etc.), in order to make it appear that the debate is ongoing and there is substantial doubt. There isn't. Like a sister comment pointed out, there is more consensus on climate change than on evolution. I expect you to wine about being downvoted and "silenced" as the next move of this playbook

1. https://skepticalscience.com/argument.php

I’m pointing out the papers in the linked study literally have different definitions.

Highlighting how this stuff really isn’t credible. It’s junk.

Regarding your link, those counter points are... very unscientific or well reasoned. Take point #1, we all agree climate changes, it snows half the year in the northern hemisphere. What caused the snow isn’t anything we did. But I’m 100% sure you could find correlations, which is all these papers do. Just like humans likely didn’t cause the earth to heat 20,000 years ago when the North American ice sheet melted. But that correlates to when humans arrived in North America.

The problem with correlations are that you need a large volume of data to suggest causation. We don’t have that.

Because of those claims of causation, the field has lost significant credibility

> The problem with correlations are that you need a large volume of data to suggest causation. We don’t have that.

Just based on this sentence I have a hard time believing you really are a scientist as you claim. There is no requirement on large volume of data to suggest causation. In fact no amount of data and correlation by itself can establish causation, because they are very different concepts (Hence the correlation!=causation). You need mechanisms to explain the causations, and you need to verify the mechanisms. This has been done extensively for climate change.

> Because of those claims of causation, the field has lost significant credibility

I would say your claims lost you a lot of credibility just now.

That's fair and I understand that. My point, is we have trouble even suggesting causation. We have one event (industrialization) and one suspected effect (warming temps). Without another time-series of data for comparison we don't know what the natural state of the world even is given the margin of error (we're only talking 1-2 degrees C).

We don't have the full dataset so effectively we're seeing two lines go up at the same time and claiming one is causing the other.

We can't even really make a correlation case, besides to say these two time series loosely correlate.

Put another way, we suspect one event causes another, which are tracked via a time series. Normally, when you're discussing causality you'd looking for multiple instances of said event causing another. We don't have that. Hence, we're effectively shooting in the dark.

I think seasons aren't really a relevant example of climate changing. Here's the Wikipedia definition of climate: "Climate is the long-term pattern of weather in an area, typically averaged over a period of 30 years." [1]

You do have a point on credibility though, in that over-politicization causes the field to look much less credible.

[1]: https://en.wikipedia.org/w/index.php?title=Climate&oldid=104...

I was simply trying to illustrate -- assume we didn't know about earths tilt; one would suspect something like the moon or stars causes the seasons (as they did historically).

IMO as we don't have accurate data measuring far enough back we're effectively shooting in the dark.

It's fine to say, these two things could be correlated, it's another to say these things are correlated AND to fix it we have to force everyone to stop traveling, be less healthy (eating a worse diet), etc.

You are creating an impossible burden of proof (causation, which in the Pearlian sense can maybe only be established when observing interventional data, which we can't have) while ignoring the balance of probabilities and possible harm done. This comic illustrates the scale of previous deviations, including the ice age one you keep harping on about

https://xkcd.com/1732/

If you drink a bunch of methanol and then go blind without knowing the exact biochemistry, you don't say "eh, there's no causation proven yet" and keep drinking, you stop the things which correlate until you can show they are safe. Where's your credibility? What have you done to give an alternative explanation to anthropogenic climate change? If you don't have one, you aren't acting in good faith here

This is a very weak response of the type that creates climate change skeptics.

1. Numbering your opponents arguments isn't a response to them.

2. He isn't asking questions, he's making a variety of statements that have a factual basis and can be debated to determine to what extent they're right or wrong. You're straw-manning to claim these are questions.

3. Asserting there's no debate when someone is telling you there is debate, is exactly the kind of strategy that makes neutral onlookers more likely to agree with him, not you, especially as one of his claims is that journals etc block any alternative viewpoint.

1. The point of numbering them is to highlight all of them have been answered already, but he's not interested in the answers. Otherwise, a good faith skeptic would have found the anders to his "concerns"

2. The debate you are alluding to has been done, see point 1. But he ignores this, because as long as the debate don't stop, his position is unchanged

3. Asserting there is a debate until someone who has done 0 to engage with an overwhelming scientific consensus agrees there isn't is false neutrality. That one random guys opinion (and mine, for the record) is worth jack shit compared to the IPCC ans the 97-99% of experts who have come to a consensus based on the data they spend their whole career studying. To go against their consensus is perfectly valid but needs to be done in a manner which integrates all the data collected in the consensus explanation at least as well, or better. Not just by listing context free gotchas and claiming the debate hasn't been finished yet

As for the question of creating climate skeptics: I'm much more interested in pointing out to readers that a rhetorical playbook is being used and provide a counter voice than to argue with someone who isn't themselves arguing in food faith. See point 1

The website you give that numbers the arguments provides answers to each one, but some of these answers are so poor as to effectively be doing work for the other side.

For example, point 43 is "CO2 is plant food". No such point is made anywhere in the post you're replying to, so you must be reading it into something else he said.

Point 44 is "there is no empirical evidence". Again, the person you're responding to didn't say that. This time I can't even see anywhere you might be reading an implication of it.

Regardless, the given answers are often not really answers at all; e.g. point 90 "Q: Peer review process was corrupted" is answered by "A: An Independent Review concluded the CRU's actions were normal and didn't threaten the integrity of peer review". I recall reading those emails and they were very explicitly doing exactly that. The "independent review" in question is actually organized and the website is hosted by the University of East Anglia itself. The answer is basically saying everyone should trust academics when they clear themselves of wrongdoing, and if you don't you're a conspiracy theorist.

I looked into this one a bit more and there's a response written by the climate skeptics themselves to this review here, which is far more convincing:

https://www.rossmckitrick.com/uploads/4/8/0/8/4808045/climat...

Russell's inquiry was slated by MPs, apparently for good reasons:

"The Muir Russell inquiry held no public hearings and did not interview any CRU critics ... Although Muir Russell had been appointed in December 2009 and was due to report in spring 2010, as of the start of April, nobody at CRU had been interviewed on anything to do with the Hockey Stick or IPCC"

An MP said: "In a situation that is almost beyond parody Muir Russell stated that he didn’t ask Jones whether he had deleted the e-mails because they would have had to interview Jones under caution. What was the solution then? The Vice Chancellor asked Jones whether he had deleted the e-mails. This rather negated the purpose of having an independent Inquiry when the only person to ask the crucial question was the Vice Chancellor who saw his prime responsibility to the good name of the University. The accused investigating themselves again."

"Subsequently, it turned out that Vice Chancellor Action’s reassurances that nothing had been deleted were untrue. Following up on that supposed reassurance, one of us (McIntyre) submitted an FOI request for the attachments to the emails which Jones had sought to destroy, attachments which contained the surreptitious and hidden edits to the IPCC report by Mann’s associates after close of external review. The University reported that the documents about which Acton had reassured the Committee did not exist after all."

enumerating numbered passages from codified documents as a response in a discussion comes across as very religious. taking it a step further and assuming bad faith and acting in accordance with a "playbook" though, that's something else entirely.

> not through discussion but rather through experimental results and verified data

Like the LHC or einstein's predictions, consensus was built for them when after high precision predictions of their models became observed in the real world. That established that correlation was causation. Climate scientists can convince skeptics if they make such detailed predictions. But to be fair, the scientific message is muffled by popular scaremongering.

Climate science suffers from the same problem as weather science, the models are essentially giant monte carlo simulations. i.e the results aren't precise.

But that is ok, because precision isn't and shouldn't be the goal. What is important is the direction, we need to know whether or not heating is occuring, roughly the rate of change and if there is anything we can to do stop it.

None of these require hard "X *C by Y year" and not being able to make such predictions shouldn't invalidate the rest of the evidence that says it's happening.

Climate science has made many predictions with a lot of success. Read the IPCC reports - read the older ones if you want to see the older predictions.

With global warming, the problem is that if you wait for the 6° warming by 2100 predictions to come true, billions are already dead.

Let’s not be hyperbolic. Without even getting into the futility of predicting something as complex as future mortality rates, the worst case projections go into tens of millions additional deaths due to climate change over the century (if that sounds like a lot, global deaths per _year_ are on the order of 50 million, so this is a few percent increase).

https://impactlab.org/research/valuing-the-global-mortality-...

https://www.nature.com/articles/s41467-021-24487-w

I expect many wars caused by famine and mass migrations. India and Pakistan will fight over water and they both have nuclear weapons.

It's very well possible to have the right conclusion ('climate change is man made') while working with imprecise data and models. High precision is not required for consensus about cause and effect to be established, merely that within the boundary of error no other explanation is viable.

They have.

Exxon's 1982 predictions about the global average temperature in 2022 are pretty spot on.

https://xkcd.com/2500/

I'm afraid that is not a good argument. There are many historical predictions that didn't come true [1]. In 1970s, there were concerns about extensive glaciation: https://en.wikipedia.org/wiki/Global_cooling. I think it is fairly uncontroversial that it is very hard to make detailed predictions about climate. This Bloomberg article discusses a recent relevant issue: https://archive.ph/FMOn5

I think a better argument would be to say that we certainly know more about climate now, and we have a lot more computational power to construct more detailed models.

[1]: There is a positive side to wrong historical predictions: that people actually dared to make concrete assertions, instead of waving hands.

> Some press reports in the 1970s speculated about continued cooling; these did not accurately reflect the scientific literature of the time, which was generally more concerned with warming from an enhanced greenhouse effect.

Well, unfortunately the lead line from the Wikipedia article is a bit misleading, as it makes it sound like it was merely a bunch of sensationalist news stories. That is not true. There were genuine concerns by climate scientists about that (apparently they were misled in part becauase of low time resolution of paleoceanography studies). Check out the references and the rest of the article.

It is probably true that there was no consensus about it back then though. (I do not know enough about it to be completely sure.)

Thank you.

But in reality science is 100% consensus. Nothing else.

There isn’t a single scientific “fact” that is anything more than high degrees of consensus.

The commenters who seem to believe that consensus is anti-science have no clue what they are talking about.

Consensus is just people agreeing on something. Science depends on tests passing, which does not depend on anyone agreeing on their results. Indeed, the point of tests is to remove the human from the 'equation' so to speak, therefore consensus isn't even possible.

If consensus is needed, it's probably because rigorous science cannot be done for whatever reason, like maybe not having enough fossil evidence.

not really , science is about building 100% bulletproof but falsifiable theories that are constantly being attacked by non-consenters and still win. No such thing as "fixed" scientific truth

Reply on news.ycombinator.com