I guess a handful of shoddy studies that are both terrible at isolating ethics and terrible at identifying social class are move persuasive than homicide, assault, and property crime statistics.
I can agree that poor people break the law more often. Not "are convicted of" but the acts.
I think the reasons are obvious and unrelated to the point of the study. Desperation, culture (broken window theory) which is critically related to average wealth, lack of opportunity, and legal veracity explains the statistics sufficiently.
Rich people have more capacity to do harm. For example that dude Luigi killed was a mass murderer, he killed 1000s, maybe 10,000s of people. So it’s hard for a poor criminal to compete
There is a finetune Qwen3.6-27B-Fable-Fus-711-UnHeretic-NM-DAU-NEO-MAX-NEO which seems to be as good at coding as vanilla Qwen, but way, way better at creative writing than Qwen and even better than Gemma 4 26 and 31b.
I saw that one in the "Popular models" sort at Hugging Face and tried it on some tasks I do frequently to compare models, and it feels damaged by the fine-tune, to me. It wrote security bugs into the code (probably just sloppy thinking, not intentional), it exhibited looping behavior in some configurations in llama.cpp, configurations I regularly use with the regular 27B, and it failed to write unit tests without being prompted (though the regular Qwen 27B tends to do so if it sees there are unit tests for everything in the repo). They have good benchmarks, but I'm not going to trust it. Also, that name is absolutely crazy.
> but way, way better at creative writing than Qwen and even better than Gemma 4 26 and 31b.
I suspect this is the only use-case I would consider...and I don't really have a use-case for "creative writing" that I would delegate to an LLM. I suppose for dialogue generation in games?
But yes, hard agree. Why on Earth would you ever want to write code with a model that is supposedly "jailbroken"? So it can put great backdoors into everything it touches? Pass.
I've noticed most fine-tunes, whether "heretic" models or something else, tend to be over-fitting, or something, at least some of the time, and get kind of chaotic. I want to believe normal folks with normal resources can be involved in this stuff, as I'm working on fine-tuned specialist models as we speak, but it seems like it takes notable investment and time. My first experiment was teaching a little Gemma 4 more to write more like me with a LoRA (like you, I don't want to use a model to write for me, but I did want training data that I could ethically use, and I've written several million words on the internet over the years), and it wasn't what I would call a success. It either wrote like an asshole (which I only do, like, 15% of the time) or it just borrowed a few of my quirks, like too many ellipses, if I applied it less heavily.
I use it to write short sci-fi stories. Life is not only about being an SDE.
> Why on Earth would you ever want to write code with a model that is supposedly "jailbroken"?
I need a model which decent enough at coding to write boilerplate for me, yet I do not need to unload in case I want to write an article or enguage in my writing hobby.Exactly why Gemma 4 are so popular.
> So it can put great backdoors into everything it touches?
You could be an Iranian sponsored bot. I'm not saying you are. You could be so don't get mad at me for publishing that statement. Because if I say "can," then I don't need to be accountable for any misinformation.
I could be! And in science when you posit a hypothesis you then back it up with data points of statistical significance.
The authors of this study have done that here. To borrow from your example, if you saw a statistically significant amount of my posts highlighting the merits of the Iranian government in a ways that run counter to the general global consensus of their actions, you would then have something that other people might agree was worth looking into. A hypothesis is not “misinformation”.
This article has not claimed to have proven anything other than outcomes in a process. I don’t understand why this is so upsetting to you.
> That's at the state level. But that doesn't account for the explosion in admin salaries and positions. The actual money a district spends on each student has been going down every year. Those funds are going more towards admin activities.
Ok? Seems like that's more of a problem than the funding. Or whatever is causing that is more of a problem, but it does a disservice to the general argument of "kids aren't receiving the same level of care" argument to blame a drop in funding--especially when it was so easily falsified.
He was asking about the trend of him commenting with "Source?" leading to downvotes on other posts. I was giving an extreme example to represent why that may be happening elsewhere. I agree that in this case with the 3% rate it's likely not applicable.
What I'm talking about is that is still considered an external capital raise for the purpose of the markets and where those assets sit on the balance sheet.
Also, keep in mind the Alphabet doesn't fully own Waymo. I don't know the percentage ownership of hand, but that also feels like it's probably a prorated investment based on ownership so Alphabet doesn't reduce its voting control.
Yes and what matters the most is what Waymo has been signaling for years. They don’t want the capex (owning and running the physical cars). I don’t know the intent of this raise but you have to realize companies may have a good asset but they don’t want to own it 100% for a multitude of reasons. Some of them could be as simple as wanting to get other investors involved and comfortable with the asset to maybe take on larger roles in future rounds. Or in this case potentially running the car part of the business.
If they truly wanted the capex, this would not be a mixed round
A fully internal recap would have been simpler. The presence of outside capital, even minority, is consistent with a gradual transition toward shared ownership, asset light structures, or operator partners.
They have made many comments over the years about this too.
Notice I left a list of potential reasons. Not that ownership has changed. Just pointing out for folks like yourself that Google has made commentary about this exploring the idea of partnering with companies that operate the physical fleet. $3bn even if chump change for you is still a larger placement and has some level of signaling indicating the want to get other folks involved at some level.
I didn't ask for potential reasons. You're talking about the "reasons" for a "gradual transition," and I am telling you that this investment isn't transitioning anything. Everyone is keeping their equal share of the company. So, I don't understand why you are giving reasons for something that isn't currently happening.
I think the words are going over your head sorry. I will try one more time but realize now it might be too much especially see some of your dead comments here.
I am not claiming a transition is happening in this round, so asking for evidence of one misses the point. Transition here means enabling future shifts in who owns and operates the capex, not changing the cap table today. If Alphabet wanted permanent full-stack ownership, an entirely internal recap would have been cleaner. Bringing in outside capital, even minority, is about signaling and optionality, not dilution.
I understand everything you have said. The D-K of you WSB transplants is wildly frustrating.
If you'll notice, all I am doing is asking the brigade of snarky know-nothings to stop talking. I'm not pretentiously claiming to know, unlike all of you. You clearly aren't in any position to understand the internal working of Google, and it's unfortunate that HN used to be a place where a question like the original one would have been answered by a person who does, but is now flooded with people like you. I will gladly take the downvotes if they're from a bunch of garage band stock pickers.
Go take a breath and stop digging a hole. Nobody is being rude to you but you are highly inflammatory and honestly a real lowering of quality. Take a bit of your medicine and step away. I am sorry you feel the need to be so rude back to everyone.
You are not “just asking questions.” You are dismissing any analysis that is not insider gossip as illegitimate, which is a convenient way to avoid engaging with the substance. No one claimed NDA level insight. We are talking about incentives, capital structure, and signaling, which is literally what outsiders analyze. If only Googlers are allowed to reason about Google, then HN has no purpose beyond rumor laundering.
I definitely modulated my tone to match yours and some of your killed comments. Sorry you don’t like what you see. Happy to have a discussion but not be told I am someone from Reddit. Low effort and low class. You are consistently being rude and you just need to reflect on some of your comments. Your right my comment back to you was definitely not nice but look at some of your killed comments. Ick.
Sure thing boss. What other advice stemming from your vast experience, wealth, and psychological heath do you have for me? I'd love to improve my image to you... the "ick" police.
My finance people care about the cents, a ROI of 7% is average but at 8.5% and now you are a world class asset of that inventory type. That’s sometimes the difference of a few hundred k out of 20m but they would not take the deal if it is slightly over due to their risk appetite.
The 3b external either matters a ton to fit their risk models OR they are doing a favor to an outside party. Probably a bit of both.
Well, given that it is an equity sale, split still feels like it is the prorated amount so that alphabet continues to own its percentage - not more not less.
Obviously you're entitled to your view, but I don't think it's that kind of finance model right now - it's far too speculative and the upside too unknown to be adjusting for small amounts on risk models.
He's talking specifically about Waymo's situation. Alplabet, a company who has $75bn of FCF, owns 80% of Waymo. A $16bn capital injection is meaningless to Alphabet, so he's wondering why they're going through the trouble.
He raises a good point, and the answer is likely that they can run into legal issues by either under or overvaluing the company in a capital raise where they're the controlling shareholder, then the IRS or existing investors have grounds for a lawsuit (or audit). They likely just want to bring the capital raise out in the open to get a fair market value, and then they will be 90% of the capital in the raise.
Hacker News likes to keep conversations focused on the topic at hand. I doubt anyone here thinks politics are irrelevant. We just understand basic courtesy. If your goal is indeed to influence change, you do a massive disservice to the cause by acting immature and injecting your politics into other conversations.
> Quite likely this year we will have federal law that will allow selling cars with fully unsupervised self-driving, in which case the insurance/liability will obviously land on the maker of the system, not person present in the car.
This is news to me. This context seems important to understanding Tesla's decision to stop selling FSD. If they're on the hook for insurance, then they will need to dynamically adjust what they charge to reflect insurance costs.
reply