And that’s why I was happier when basic research was mostly being carried out in public institutions.
For example, if you join a Math department and start publishing papers that cast the department's research interests in a negative light, that's often a good way to get fired.
I mean, seriously, go ask just about any professor of any topic at any university if their field is undervalued or overvalued. They'll almost universally tell you that their field is undervalued, underappreciated, and more important than people think. They deserve more funding, and you should definitely consider majoring in their field.
I find it difficult to come up with an analogy from a public institution. What is the product the math department would be pressuring it's members to protect?
Sure, each department in a university believes it's work important. That doesn't seem even remotely similar to me as this issue with Google.
> It's not theoretical, researchers are quoted in the article and papers have been altered.
In grad school, PhD students' advisors typically insist on various revisions before publishing a paper, as publications reflect on the advisor and their institution. So there's nothing even slightly weird about Google having the same interest in revising the papers pushed by its researchers.
Unless there IS something weird about what Google's doing? But if so, what is it?
According to this Reuters article, Google's new process happens after peer review and Google's other processes have completed.
"The “sensitive topics” process adds a round of scrutiny to Google’s standard review of papers for pitfalls such as disclosing of trade secrets, eight current and former employees said."
Instead of publishing the paper, Google will now review the paper with an eye to negative impact the paper may have on existing Google product (or lobbying efforts, etc.) Google isn't doing this to improve the quality of the paper, they are doing it to protect their business interests.
"For some projects, Google officials have intervened in later stages. A senior Google manager reviewing a study on content recommendation technology shortly before publication this summer told authors to “take great care to strike a positive tone,” according to internal correspondence read to Reuters."
I think this is very different from the advising process in graduate school.
Preventing people from disclosing trade secrets seem fair to me. Preventing valid research simply because it may negatively impact business strikes me as less reasonable.
"Four staff researchers, including senior scientist Margaret Mitchell, said they believe Google is starting to interfere with crucial studies of potential technology harms."
All of that sounds normal to me. Including filtering out trade-secrets, which is completely normal when working with trade-secrets in grad school too. Additionally, it's completely normal to filter out intellectual property you might plan to patent; confidential information; proprietary industrial information; information protected by law; dangerous findings (e.g., hackers often omit details of an exploit until the relevant vendor has had time to fix); and a few other categories.
Maintaining a positive, constructive tone is also completely normal. For example, failed experiments are typically described as progressive steps toward an ultimate success; unforeseen problems are discoveries; and major issues are seen as research challenges to be overcome. Or, ya know, stuff like that.
I mean, is that all this story's about? Because if that's it, then it seems like nothing substantial. But if that's the case, why is this in the news?
couldn't the exact thing that happened at Google happened at a university? For example one researcher could publish a paper criticizing the methods that other researchers in the department have developed because of their carbon impact.
I think it's fair to say that there are internal pressures not to do that - - such a professor would have a hard time thriving in the department if they are attacking the methods of their colleagues.
The Gebru paper makes a lot of hay about carbon impact, but Google's put a ton of effort into mitigating that, they've been carbon neutral on paper since 2007: https://blog.google/outreach-initiatives/sustainability/our-...
Do you have an example of a public university censoring research papers that passed peer review because that university believed the paper cast "an innacurate negative light"... on what, exactly?
My rejoinder was that if it's inaccurately critical of Google, like hyping carbon impact without mentioning a decades-long carbon mitigation program, I get a lot more sympathetic to Google's position. Why should they pay someone to spread falsehoods about them?
I don't really care about the process minutiae.
"Large language models can have an outsized carbon emissions impact, but not Google's, those are carbon neutral".
If you're going to bring carbon emissions into it, mentioning current mitigations as well seems relevant unless you're writing an op/ed piece.
The "mitigation" in question is buying carbon offsets (I mean there are improvements in DC efficiency also, but those only do so much, and language models ballooning 100x isn't going to be fixed with 10 or 50% efficiency improvements). For the moment "carbon neutrality" is only achieved through the purchase of energy offsets.
That doesn't mitigate. It offsets. Don't get me wrong, still better than nothing, but its not a mitigation.
* this generation of language models leaning into transfer learning reducing the total number of training runs for different applications
* TPUs being more power efficient than GPUs (the numbers they used in the paper were based on GPUs)
* other energy-centric stuff that's not just offsets, efficiency like you mention in addition to sourcing from renewable
I’m not sure about math, but in physics it would be string theory which has been a dead end and has mostly served as a welfare program for boomer scientists
I still have faith in the research done by hard science research teams.
With Math, the sensitive topic is utility. Math departments would tend to be upset by someone pushing papers that, say, call attention to the lack of social benefit relative to other fields of research that taxpayers could be funding instead.
I guess a researcher could also spin a narrative about mathematicians' contributions to cryptography, examining the negative consequences enabled by stuff like Bitcoin.
---
> I still have faith in the research done by hard science research teams.
That gets kinda complicated.
I mean, hard-science teams often rely on their expertise, as it's their unique source of value: if the area that they excel in is shown to be sub-optimal, admitting it would mean losing their current career path (and often taking a significant tumble down-the-ladder after transitioning).
So hard-science teams are often largely academically honest in what they do publish, though they're often biased toward casting what they do in a positive light.
For a common example, have you ever met an older computer tech who insists on using a legacy, obsoleted technology because it's what they know best? A lot of academics basically do the same thing.
This is most likely due to an availability bias. They have concrete examples in their own sphere but not near as much detail about competing domains. This leads to the mistaken assumption that their own field more important than it possibly would be if viewed in the overarching context
Just because bias is impossible to avoid, does not mean that all bias is the same.
The goal isn't outright elimination of bias (which is completely impossible), its ensuring incentives that are aligned so bias doesn't become a self-reinforcing feedback loop.