What does this even mean? Every initiative is kept alive by money. It was very clearly attacked by the Trump administration, and once they started attacking, DEI initiatives stopped. Why else do you think the "implosions", apparently happening all the time, were finally too much to sustain around 2025?
When did DEI go too far, and why? I am not aware of any anti-DEI position that thinks it was ever an acceptable practice. It has been resisted since the 80s.
Seems like there is obviously a spectrum of positions here and various people have found themselves calling certain positions "DEI" or "anti-DEI" or "affirmative action" or "racist" depending on which positions they hold and where a particular institution lies on it. For example:
1. A company that hires mainly based on on-paper qualifications (which, due to existing disparities in experience and education, means in practice that they hire fewer historically disadvantaged minorities)
2. A company that, believing that talent or merit is more equitably distributed
among groups than on-paper qualifications may suggest, goes out of their way to interview a greater proportion of historically disadvantaged minority groups than necessarily would naturally apply to their positions (to discover those as-of-yet-unrecognized talents), and tries to hire based on actual merit and ability rather than on-paper qualifications.
3. A company that has the goal of creating a more equitable and diverse workforce with groups that more closely matches their proportions of the population. They spend a lot of time and effort actively recruiting and seeking out employees from underrepresented minority groups, meaning that it is effectively easier for minority groups to get interviews at this company for this position, but still tries to hire based on actual ability and talent once they get to the interview stage
4. A company which, accounting for the disparities in distribution of measurable abilities of groups, tries to hire based on within-group ability rather than between-group ability; for example trying to hire the top 10th percentile of black, or top 10th percentile asian, or top 10th percentile male, rather than top 10th percentile of all applicants in aggregate (which would likely be overrepresented by over-performing groups). This in practice means using a different standard for different groups, but using the within group distribution of talent; they may say they are "considering race/sex/disability as one of many factors they take into account"
5. A company that in practice has an explicit hiring quota based on the race/sex/etc. of applicants, in order to create a more equitable or diverse workforce that matches the relative sizes of those groups in the local population.
6. A company that, in order to reverse historical and current disparities in achievement within society, actively discriminates in the opposite direction as historical trend, with the goal of evening the playing field more quickly.
7. A company that, in order to address historical grievances and grant some amount of justice to victims of discrimination and oppression, actively and purposefully discriminates against privileged majority and overachieving groups, favoring underprivileged and underachieving groups, up to and including reparations for past injustices to remedy current disparities.
I'm pretty sure 1-3 are legal and 4-7 are illegal, but I'm not a lawyer so that's just my guess. Some hardliners might attack 2 as going too far, but somewhere between 3 and 4 is probably the more common point at which anti-DEI people draw their line.
The history is more than a little complicated. I model it as a set of separate overton windows in different institutions that have shifted, widened, and contracted over the last 60 or so years. So in this picture, courts arguably expanded it to 4-6 in the 80s, but have retreated back to around 2-3 since then. Legislature has mostly stayed around 2-3 with a handful of radicals pushing higher or lower with low-to-moderate success. Academia has been around 5-7 since the 70s. Executive has mostly hovered around 3-5 until Trump. Media/popular culture, big shift up to 6-7 in 2014-2021, retreating back to 3-4 now. Also, motte-and-baileys are common here; pro-DEI activists sometimes frame it as something like 2, but often advocate up to 6 or 7 (e.g. "antiracism" in the 2020s peak). anti-DEI activists often say they simply oppose quotas and illegal discrimination (4-7), but in practice sometimes attack even 2-3 as being too much.
> They spend a lot of time and effort actively recruiting and seeking out employees from underrepresented minority groups, meaning that it is effectively easier for minority groups to get interviews at this company for this position, but still tries to hire based on actual ability and talent once they get to the interview stage
Discriminating against candidates on the basis of protected class at the application review stage is no less illegal than discriminating during interviews.
#2 may or many not be illegal, depending what you mean by, "goes out of their way to interview a greater proportion of historically disadvantaged minority groups than necessarily would naturally apply to their positions." If this involves using protected class as a factor, in any way, on which candidates to advance to interviews, then yes it is illegal discrimination.
part of the issue is that there are multiple laws and legal frameworks even at the federal level in the US that are kind of contradictory here. See "disparate impact" law, where using a neutral standard (such as a background check, or an ability test) that affects different protected groups differently can be considered illegal discrimination. 2 might be illegal, but 1 might be illegal too in the other direction! It kind of seems like the only solution is to basically maintain plausible deniability and try your hardest not to leave an auditable paper trail about any of the reasoning you use to hire a candidate or not.
Fear of disparate impact claims do not justify discrimination on the basis of protect class. This was argued all the way up to the Supreme Court: https://en.wikipedia.org/wiki/Ricci_v._DeStefano
Furthermore, disparate impact theory has been largely rolled back in recent years.
The Sheetz disparate impact case was in 2024, post Ricci. Griggs is still the law of the land even though some justices have signaled they might want to revisit it.
Ahh, the imaginary butthurt of those for whom “discrimination” has only ever been a shorthand for “having to compete on a level playing field”.
Trump (and his ilk) presented a straw man to the dumb, then shot the straw man, without ever thinking of the innocent people standing down range of their target, which included the dumb the whole time.
The same administration who, just yesterday hired a new press secretary who was very obviously not hired for qualifications but instead for being a blonde.
Doesn't seem like that's implied at all. There is no implication about whether it is good or bad actually, but if the fact it was attacked is a bad thing, the only inference is that on balance it did more harm than good.
I want to agree with you, but most software projects people work on are ridden with unknown bugs and people do in fact live with this. I'm not aware of any commercial software project that arrives at the goal "not having any bugs or unforeseen failures at all". And where this does happen (space probes?) it is very expensive.
IMO the issue with LLMs is they massively widen the "unknown unknowns" (skipping review makes this problem even worse). This is obviously much worse when an LLM is actually a part of the software rather than just producing software that is otherwise deterministic.
I think OP was a little too absolute but i agree with the premise. People who engineer buildings and bridges are not allowed (generally) to ship mistakes. Software "engineers," though, have just accepted it as standard practice.
I do think the world would be better if commercial software had to meet some professional standard of fitness, with meaningful repercussions if not. I also realize that ship apparently sailed before i was born, and I'm not paid to do serious engineering, I'm paid to ship features.
(Not even saying I'm qualified to do the former anyway, i just also think every time national air traffic is shut down or like in France the other day and hundreds of supermarkets lose all their frozen inventory, or i can't do my job because some service failed, or yes, claude code is slow and heavy age can't work right in every terminal, that maybe we shouldn't be writing the stuff that the whole world relies on to function every day with zero agreed upon professional standards)
Exactly! That's my point. It's already a problem, and it's not "fine" just because it's the status quo, but at least we can chalk it up to "humans are fallible". LLMs 100x the amount of mistakes a programmer can make in a day, and that's inexcusable to not take seriously, that's my point. We don't need to chase perfection, but we can't pretend like whatever Claude one-shots is any good at all.
> LLMs 100x the amount of mistakes a programmer can make in a day
1. Have you seen that? If so, which agent+model?
2. Is “per day” intentionally used instead of per implementation - i.e a coding agent may A) be far faster tok/s than a human, and B) work more hours per day, so it’s similar bug rate per implementation? Or did you mean bugs per implementation?
1. Are you referring to LLMs handling 100x as many patches as a human can, or that LLM implementations are 100x as long as a human's implementation?
2. In my experience, LLM implementations are often 2-4x as long, largely because they have a lot more error handling than I would have chosen to implement. However, I don't find they have correspondingly X times as many mistakes.
3. Part of how agents seem to avoid many mistakes or even catch mistakes is by testing things. For example, it caught issues in Pester functions on certain Pester versions, or will even analyze binaries to verify the right config settings to use, etc.
4. If you are seeing 100x as many bugs, which agent+model?
LLM is clearly not 100xing the mistakes as we have witnessed Anthropic’s success. Otherwise the system would not have been functional. It’s not like their system is not tested.
The thread is about jankiness, lack of feature cohesion and code quality.
What we are discussing here is: is it acceptable to accept poor code quality and trade it off with delivery?
> What we are discussing here is: is it acceptable to accept poor code quality and trade it off with delivery?
I found that people that lean towards the latter also favor profits over customer satisfaction. Then they will try to defend themselves with a black and white view “it’s either this scrap or no delivery”.
> I'm not aware of any commercial software project that arrives at the goal "not having any bugs or unforeseen failures at all".
It's ludicrous that we have executives at tech companies that think this is possible.
One S&P 500 company I worked for signed a big contract to deliver bugless software for an enterprise web portal for the state of Minnesota. I was on the team that had to make that magic happen. I had to leave the company not long after that due to nearly life-ending burnout.
So I would throw scorn at the ivory tower inhabitants here who should know better than to suggest that anything of substantial complexity can be bug free
It may not be necessary to think in those terms exactly, but if you are not even able to think in those terms, you probably will be no good at prompting the LLM either. It's quite plausible that someone can be a productive dev with claude if they don't know exactly how the message queue is processed. But if they don't know there is a message queue at all, that is much less likely.
Likewise, you don't care exactly how an if statement gets converted into machine code, but you do know precisely what an if statement is and how it should behave, and could identify if it was buggy, and that that part of the codebase contains a bug. If you can't do that, then there is an impossible-to-estimate probability that at some point you get stuck and no progress will ever be possible. I don't see that as a winning strategy, in the long run (but it may work very well in the short term).
What do you think "steering the AI" means? It is not like we need someone to sit at a desk and type "yes, implement it". Of course if you are a sane person you mean someone who will determine if the AI is doing "the right thing" and change its direction if it is wrong. That person is almost by definition a senior engineer.
Someone making sure that the business requirements are implemented properly and correctly. Basically a project manager. Maybe also someone testing the output and providing feedback. Not someone that says `we might have a race condition here, lets implement a distributed lock`.
I think that person does not need to know about locks anymore.
At the moment yes, you are absolutely right. The argument was about a couple years into the future. Nobody can predict this though. I feel this will slowly steer that way. We’ll live and see.
> No one is disputing that the Lean formalization of Navier-Stokes is correct, so we should have high confidence that the generated Lean proof is valid.
Why? It is a correct proof. Is it a correct (dis-)proof of NS? These statements do not seem particularly related. People, including (especially?) mathematicians, find NL proofs far easier to read than Lean proofs. I would bet the ratio of people who read (part of) the NL proof vs the Lean proof is at least 100 to 1. It seems incredibly unlikely that anyone has read and reviewed the whole Lean proof (including the authors of the OP).
Is it particularly easy to see that the Lean proof does actually disprove NS, rather than proving something similar but not the same?
Well the theorum statement (what it proves) is short and easy to verify. I would not be surprised if hundreds of mathematicians have looked at it at this point.
It doesn't take that many people to perform rent extraction -- you can have an entire country run by one absentee landlord so long as you can pay the local police to maintain order for you. Obviously the counterfactual is not available to us, but the colonial administration presided over the deaths of tens of millions due to disease and famine. Some of that is unavoidable tragedy, but quite a lot of it is direct mismanagement (especially fervent devotion to a "work sets you free" mindset and free market capitalism).
We are indeed not counting the people doing the work. The colonial office does not include the actual administration of the country, that is kind of the whole point of my comment.
I am in the process of reading the volume on India for the Oxford History of the British Empire.
In early modern academia, "Historia" was defined as opposed to "Philosophia" as being the study of the particular versus that of generalizations.
One thing that my experience in studying history at a major European university showed me is that almost any generalization one makes about any period in history can be disproved. There is a lot of complexity in the development of path dependent evolution and irreversible processes that simply cannot be compressed and modeled.
That's comparable to the Victorian elites' situation, where they only count people directly employed by the home office, and ignore the so called independent colonial administration as if it's a completely different category of human.
Are they not also Victorian Elites, or did they lose that status by not being in Britain?
Above we are counting them. This about ~10000 people total, most of them abroad. We are not counting military since these are civil servants, but neither are we counting military today.
I feel like not counting the military is silly in that case, because the military in colonies was essentially involved in police actions, engineering duties, and ensuring administrative rules were followed.
As an essential part of the colony, you can't really count them out. That's similar to how in feudal structures, and military regimes, I'd count the military and civil service as one entity.
>Some of that is unavoidable tragedy, but quite a lot of it is direct mismanagement (especially fervent devotion to a "work sets you free" mindset and free market capitalism).
It was all going so well until you spewed propaganda, believe it not the provinces administered by the British were least affected by disease and famine primarily because of the infrastructure and institutions that'd been established in them.
Additionally, we have a better insight into those regions because of the record keeping again introduced by the bureaucracy forced upon them...
You just have to look at past colonies today in india to still see the positive impact they had. To this very day all proper and impressive buildings are the ones build back then, sadly they are all slowly falling apart because of lack of care and maintenance by the locals.
This is delirious exaggeration. The problem has not even been widely recognized for several years. Fable reported "two rs in raspberry" to me as recently as August. There is some randomness, it's hard to predict which words will trip up the machine, and I haven't been able to do it at all since August. But it was absolutely happening until very recently, and probably still is.
Training a model is not "learning something". Only people learn things. Whether training is a fair use is debatable, but it has nothing to do with the justification that people are allowed to learn from books.
reply