Hacker News Reader

Comment by gammarator | Hacker News Reader

Here’s an extended comment by another astrophysicst: https://telescoper.blog/2025/01/02/timescape-versus-dark-ene...

The most important bit:

> The new papers under discussion focus entirely on supernovae measurements. It must be recognized that these provide just one of the pillars supporting the standard cosmology. Over the years, many alternative models have been suggested that claim to “fix” some alleged problem with cosmology only to find that it makes other issues worse. That’s not a reason to ignore departures from the standard framework, but it is an indication that we have a huge amount of data and we’re not allowed to cherry-pick what we want.

View on HN

the thing is, this is not really an alternative model. it's rather actually bothering to do the hard math based on existing principles (GR) and existing observations, dropping the fairly convincingly invalidated assumption of large scale uniformity in the mass distribution of the universe.

if anything the standard model of cosmology should at this point be considered alternative as it introduces extra parameters that might be unnecessary.

so yeah it's one calculation. but give it time. the math is harder.

This has the same number of free parameters as LambdaCDM. Also this result only looks supernovae, i.e. low redshift sources. LambdaCDM is tested on cosmological scales.

Very interesting, but “more work is needed”.

thats not the case, if, as is increasingly speculated, the lambda is not constant over time. you figure two parameters for linear and three for a quadratic experience

Calculation is harder in a world of functionally limitless compute is sort of interesting. Where do we go from here?

> dropping the fairly convincingly invalidated assumption of large scale uniformity in the mass distribution of the universe.

The problem with that is then you need a mechanism that creates non-uniformly distributed mass.

Otherwise, you are simply invoking the anthropic principle: "The universe is the way it is because we are here."

You don’t need a mechanism to point out a fact contradicts an assumption, eg, our measurements show non-uniform mass at virtually all scales (including billions of light years). There simply is no observable scale with uniform mass.

Obviously there’s some mechanism which causes that, but the mere existence of multi-billion light year structures invalidates the modeling assumption — that assumption doesn’t correspond to reality.

yeah the ~1b ly nonuniformity is pretty much there. the ~10b ly uniformity is still early days but looking more and more likely as more data roll in (unless there is a systematic problem)

I think that can be mitigated in three ways: our understanding of inflation is flawed, there were more "nucleation" sites where our universe came to be, and there are the already theorized baryonic acoustic oscillations that could introduce heterogeneity in the universe.

Maybe is a combination of these, maybe something else. If nothing else, the uniformity is less probable than a mass distribution with variance (unless there is a phenomenon like inflation that smoothen things out, but also that was introduced to explain the assumption of a homogeneous universe). I concede that explaining the little variance in the CMB with our current understanding is hard when dropping homogeneity assumption however.

> The problem with that is then you need a mechanism that creates non-uniformly distributed mass.

The mechanism is gravity; and we have good observational evidence that the mass distribution of the universe is not uniform, at least at the scales we can observe (we can see galaxy clusters and voids).

> The problem with that is then you need a mechanism that creates non-uniformly distributed mass.

you need no such thing. thats like saying "i refuse to acknowledge the pacific ocean to be so damn large without a mechanism". you dont need that. it just is. this doesnt preclude the existence of such a mechanism. but for any (legit) science, mechanistic consideration should be strictly downstream of observation.

That sounds like regression.

If this problem of regression occurs as regularly as your quote implies then the fault is not in these proposed alternatives, or even in the likely faulty existing model, but in the gaping wide holes for testing these things quickly and objectively. That is why us dumb software guys have test automation.

I think automated hypothesis testing against new data in science is itself an incredibly difficult problem. Every experiment has its own methodology and particular interpretation, often you need to custom build models for your experimental setup to test a given hypothesis, there are lots of data cleanup and aggregation steps that don't generalise, etc. My partner is in neuroscience, for instance, and merging another lab's data into their own workflows is a whole project unto itself.

Test automation in the software context is comparatively trivial. Formal systems make much better guarantees than the universe.

(not to say I think it's a bad idea - it would be incredible! - but perhaps the juice isn't worth the squeeze?)

> Every experiment has its own methodology

That is bias. Bias is always an implicit default in any initiative and requires a deliberate concerted effort to identify.

None of what you said is unique to any form of science or engineering. Perhaps the only thing about this unique to this field of science, as well as microbiology, is the shear size and diversity of the data.

From an objective perspective test automation is not more or less trivial to any given subject. The triviality of testing is directly determined by the tests written and their quality (speed and reproducibility).

The juice is always worth the squeeze. Its a business problem that can be answered with math in consideration of risk, velocity, and confidence.

Respectfully I disagree - the situation is far more complex in science than software engineering disciplines.

I agree that different tests require different amounts of effort (obviously), but even the simplest "unit tests" you could conceive of for scientific domains are very complex, as there's no standard (or even unique) way to translate a scientific problem into a formally checkable system. Theories are frameworks within which experiments can be judged, but this is rarely unambiguous, and often requires a great deal of domain-specific knowledge - in analogy to programming it would be like the semantics of your language changing with every program you write. On the other hand, any programmer in a modern language can add useful tests to a codebase with (relatively) little effort.

We are talking hours versus months or even years here!

The experiment informs the ontology which informs the experiment. I don't think this is reducible to bias, although that certainly exists. Rather to me it's inherent uncertainty in the domain that experiments seek to address.

Business practice, as you use the term, evolved to serve very different needs. Automated testing is useful for building software, but that effort may be better spent in science developing new experiments and hypotheses. It's very much an open problem whether the juice is worth the squeeze - in fact the lack of such efforts is (weak) evidence that it might not be. Scientists are not stupid.

> We are talking hours versus months or even years here!

That is why there are engineers that specialize in how to perform testing so that it doesn't take so long. For example long tests don't need to run at all if more critical short tests fail. The problems you describe for astrophysics are not unique to astrophysics even if the scale, size, and diversity of the data is so unique. Likewise, all the excuses I hear to avoid testing are the very same excuses software developers make.

The reality is that these are only 25% valid. On their face these excuses are complete garbage, but a validation cannot occur faster than the underlying system allows. If the system being tested is remarkably slow then any testing upon it will be, at best, just as remarkably slow. That is not a fault of the test or the testing, but is entirely the fault of that underlying system.

Uniqueness is not a requirement =) I still think you are over-generalising from experience in the software world. Science is about sense-making, finding parsimonious ontologies to describe the world. Software is about building reliable automation for various purposes.

They have orthogonal goals; why would you believe that automated testing would work the same way in both domains? I just don't see it.

Maybe you can elaborate on what you mean by automated testing of scientific hypotheses? I get the feeling we are talking past each other because we're both repeating the same points. Maybe we should focus on the 25% of excuses you've agreed are valid!

> I just don't see it.

That’s because you haven’t tried. You only test what you know, what’s provable. The goal isn’t 100% validation of everything. The only goal is error identification. Surely you know something in science, like the distance to Proxima Centauri. Start with what you know.

Then when something new comes along the only goal is to see which of those known things are challenged.

Testing doesn’t buy certainty. It’s more like insurance, because it successfully lowers risks in much shorter time. Like with insurance there aren’t wild expectations it’s going to prevent a house fire, but it will prevent unexpected homelessness.

You are oversimplifying science, especially theoretic physics. At the point where we are, there are neither any quick/cheap tests, and there is no objectivity. The space of possible correct theories is infinite, and humans are simply not smart enough to come up frameworks to objectively truncate the space. If we were, we would have made progress already.

There is a lot of subjectivity and art to designing good experiments, not to mention a lot of philosophical insight. I know a lot of scientists deny the role of philosophy in science, but I see all the top physicists in my fields liberally use philosophy - not philosopher type philosophy but physicist type philosophy - to guide their scientific exploration.

I am not. You are using bias as an excuse to qualify poor objectivity. I am fully aware that astrophysics contains a scale and diversity of data beyond my imagination, but that volume of data does not excuse an absence of common business practices.

> The space of possible correct theories is infinite

That is not unique to any form of science, engineering, or even software products.

> and humans are simply not smart enough to...

That is why test automation is a thing.

> and humans are simply not smart enough to come up frameworks to objectively truncate the space.

We are, but some people stubbornly resist such things. For instance, MOND reproducing the Tully-Fisher relation and being unexpectedly successful at making many other predictions suggests that any theories purporting to explain dark matter/gravitational anomalies should probably have MOND-like qualities in some limit. That would effectively prune the space of possible theories.

Instead, they've gone in the complete opposite direction, basically ignoring MOND and positing different matter distributions just to fit observations, while MOND, against all odds since it's not ultimately correct, continues to make successful predictions we're now seeing in JWST data.

Note that when I say humans, I mean how the whole worldwide social institution/network of scientists, with all its inherent hierarchies and politics. Because that is the social network that ultimately comes to some consensus on what the best theories at any given moment are.

Its indeed possible, in fact "necessary" that some individual scientists within this network by luck or brains come up with much better theories, but sometimes those theories are not accepted by others or they are. But ultimately all that matters is the consensus.

Holy fuck guy. Take a step back and do some self-reflection. Any time people post about physics on here its all emotions about how hard life is. With so much longing for sympathy its amazing anything in the field ever gets published.

Unless you are looking for research grants stop crying about consensus and instead return to evidence and proofs. There will always be a million sad tears in your big sad community. If that is your greatest concern its going to take you a million years to prove what you already know, because all the sad people you are showing it to are just as sad and self-loathing about social concerns as you are.

Reply on news.ycombinator.com