Hacker Newsnew | past | comments | ask | show | jobs | submit | cebert's commentslogin

I’m thankful that Meta still contributes to open source and shares models like this. I know there’s several reasons to not like the company, but actions like this are much appreciated and benefit everyone.

Does everyone forget 2023 when someone leaked the llama weights to 4chan?? Then meta started issuing takedowns on the leaks trying to stop it.

Meta took the open path because their initial foray into AI was compromised so they have been doing their best to kneecap everyone else since then.

I like the result but let’s not pretend it’s for gracious intent.


Wait a minute. I'm no Meta fan, but that leak wasn't internal. llama released their weights to researchers first. The leak was from the initial batch of users, not from inside of Meta. iirc, the model was never meant to be closed weight.

I agree How can the previous comment be on hacker news ? Every one here has followed the llama release saga. The famous cheeky PR on their GitHub with the torrent link was genius comedy.

This might make sense for explaining n=1 releases of Llama being open weight. Even OpenAI started with open weight models and moved to closed weight though, so why would this have forever locked Meta into releasing all models as open weight and across so many model families if they weren't really interested in that path as a strategy in its own right?

There is so much malice in the world, let’s just pretend for once it is gracious intent. Feels better.

Sure, but people have the right to ask questions, as for example Zuck's pledge to give away 99% which people pointed out might be a tax avoidance scheme

The retort was essentially "Can't you just be nice?" but people have the right to ask questions; sometimes the questions reveal much corruption that actually does go on


I think it is valid to question why he'd be giving away 99% of his fortune, because let's be honest, Zuck has not proven that he is trustworthy. But at the same time, he could just... Not donate that much.

Yes, the 99% did NOT go straight into non-profits, instead being funneled into his foundation, which has donated millions into actual charitable organizations, but that's arguably millions that wouldn't have otherwise gone to those orgs.

Is it a bit disingenuous to say he's donating 99% of his wealth when his foundation has only donated a few hundred million (or few billion?), which is a single percent of his wealth? Yeah, probably. But a few billion is more than zero, and is undeniably helpful to those organizations.


Not of fan of the company for the social media but have to appreciate all the open sourcing. none of the other top labs release thier models like meta.

> none of the other top labs release thier models like meta

Don't basically all the "top labs" except Anthropic now have open weight models? And Zuckerberg said they were now going to be "careful about what we choose to open source" in the future, which is a shift from their previous rhetoric about "Open Source AI is the Path Forward".


They're not doing it out of the goodness of their heart, they're deploying a classic strategy known as "Commoditize Your Complement"[1], to ward off threats from OpenAI and Anthropic. It's only a happy accident that the little guy benefits in this instance.

Facebook is a deeply scummy company[2] and their stranglehold on online advertising spend (along with Google) allows them to pour enormous funds into side bets like this.

[1] https://gwern.net/complement

[2] https://en.wikipedia.org/wiki/Careless_People


Not even closely OK with facebook. But none of the other companies do this. And Mark has been open about it. I remember him saying in an interview the same very openly. Something oddly respectable about NOT sugar coating with good PR and marketing. Unlike OpenAI.

Well, when your incentives happen to align with those of a faceless mega-corporstion, you gotta take what you can get.

You dont have to thank them for it though.

I spend years working on training these models. Inference is always the fruit. The effort going into getting the data is the most time consuming part. I am not a fan of meta from a long time. But open sourcing the weights help move the field in general. So I have to be thankful for that.

you don’t, that’s true

i prefer to say thank you when someone is doing something good


We can still like it. We're not nominating Nobel Prizes or something.

Among the top 10 tech companies and beyond, they have the most successful open source program.

These projects come to my mind:

SAM segment anything.

PyTorch

LLama

...

Open source datacenters and server blueprints.

the following instead comes from grok.com

Meta’s open-source hall of fame (Nov 2025)

---------------------

Llama family (2 → 3.3) – 2023-2025 >500k total stars · powers ~80% of models on Hugging Face Single-handedly killed the closed frontier model monopoly

---------------------

PyTorch – 2017 85k+ stars · the #1 ML framework in research TensorFlow is basically dead in academia now

---------------------

React + React Native – 2013/2015 230k + 120k stars Still the de-facto UI standard for web & mobile

---------------------

FAISS – 2017 32k stars · used literally everywhere (even inside OpenAI) The vector similarity search library

---------------------

Segment Anything (SAM 1 & 2) – 2023-2024 55k stars Revolutionized image segmentation overnight

---------------------

Open Compute Project – 2011 Entire open-source datacenter designs (servers, racks, networking, power) Google, Microsoft, Apple, and basically the whole hyperscaler industry build on OCP blueprints

---------------------

Zstandard (zstd) – 2016 Faster than gzip · now in Linux kernel, NVIDIA drivers, Cloudflare, etc. The new compression king

---------------------

Buck2 – 2023 Rust build system, 3-5× faster than Buck1 Handles Meta’s insane monorepo without dying

---------------------

Prophet – 2017 · 20k stars Go-to time-series forecasting library for business

---------------------

Hydra – 2020 · 9k stars Config management that saved the sanity of ML researchers

---------------------

Docusaurus – 2017 · 55k stars Powers docs for React, Jest, Babel, etc.

---------------------

Velox – 2022 C++ query engine · backbone of next-gen Presto/Trino

---------------------

Sapling – 2023 Git replacement that actually works at 10M+ file scale

---------------------

Meta’s GitHub org is now >3 million stars total — more than Google + Microsoft + Amazon combined.

---------------------

Bottom line: if you’re using modern AI in 2025, there’s a ~90% chance you’re running on something Meta open-sourced for free.


OSQuery

I dont think it's open source. It says SAM license. Most likely source available.

Agreed. The community orientation is great now. I had mixed feelings about them after finding and reporting a live vuln (medium-severity) back in 2005 or so.[1] I'm not really into social media but it does seem like they've changed their culture for the better.

[1] I didn't take them up on the offer to interview in the wake of that and so it will be forever known as "I've made a huge mistake."


If they really deliver a model that can track and describe existing images / videos well that would be a huge breakthrough. There are many extremely useful cases in med, law, surveillance, software and so on. Their competition sucks at this.

Disappointingly, every time Zuck hands out some free shit people instantly forget that he and his companies are a cancer upon humanity. Come on dude, "several reasons to not like the company" doesn't fucking cut it.

I would love to know what the actual new terms are, such as the new carriage fee YouTubeTV pays.

Good idea. I was hoping to at least see an overview of this from my phone, but when I opened the link, it said it’s for desktop only and became uninterested.

Just pushed an update for mobile - should work now. Happy to give you a free license key if you want to try it on your Mac. Would love feedback

You can easily use the same types and libraries in your backend and frontend with TypeScript. It’s not at easy with dotnet.

While suffering the performance loss of V8 versus CLR, JVM or any compiled language.

One of the reasons I am back to writing more C++ code is C++ addons for node.js, as several SaaS products now only care about Next.js as extension SDK.


> While suffering the performance loss of V8 versus CLR, JVM or any compiled language

The number of startups for whom that performance differential matters more than developer output is tiny.


Yeah, except plenty of them are probably using Kubernetes and NoSQL, because everyone dreams to be Google.

I don't think it's about dreaming to be Google. K8s is pretty easy to set up now with a hosted cloud platform if you start with it, and helm takes care of pretty much all your infra needs. Migrating to K8s is what's awful. From there, the docs have most everything you need to know and there's an abundance of helpful information online that covers most problems you'll run into.

>You can easily use the same types and libraries in your backend and frontend with TypeScript. It’s not at easy with dotnet.

You can do that in .NET, too if you use Blazor for frontend.


OpenAPI and client generators solve this issue easily.

I would love this to be true, but it isn't. I've done generating types for the frontend multiple times, sometimes from C# (around 2016, using typelite), Java (openapi template generator) and most recently straight from OpenAPI spec files (.yaml) using Orval.

It always has been a shitshow. It works well for the 90% cases, but in the 10% edge cases, things break. It becomes impossible to fix generation issues, you will often resort in working around issues in your backend/openapi code. Sometimes you report bugs upstream and hope it gets fixed. In the current project we are stuck on a ~2year old Orval version (a typescript generator from openapi) because some features broke or were removed in the latest version, and the entire monorepo (15+ LoB apps) wouldn't compile and would require major changes. This simply because a never version of the generator was broken/removed features previously present.


No, that's not true. If you share code like this then you can do things like put the same validation code in the frontend and the backend: frontend to give a nice user experience, and backend to protect the endpoint.

OpenAPI does support patterns for fields and nullables/non-nullables - that already gets you very far regarding validation. A decently sophisticated generator (which don't exist IMHO) would generate the validation code for your respective language.

True, but you can get all the way to zero duplication if you write it directly and share that code.

Or GraphQL.

Still one lang on both ends is nice: there are some bits of code you want to run on both ends (like templating for SSR/SEO/caching; but also using them in the browser).


Still more work than just running the same code everywhere.

Still better than bringing JavaScript to the back end shudders

If this can’t support web standards it’s a nonstarter for me.

That’s not exactly fast for people who need these tools though.

I love the suggestion here of having a tool that can search for other tools to preserve context.


Does anyone have something positive to share about UnitedHealth?


Yeah I really like their rewards program: https://e-i.uhc.com/uhcrewards

Each year they pay me $1,000 (in the form of HSA deposits, which I can invest) to do basic things like get a checkup, get a flu shot, and get a blood test. I sync my wear-able data and they pay me $1-2 each time I exercise or get enough sleep.

It's great!


>I sync my wear-able data and they pay me $1-2 each time I exercise or get enough sleep

I wonder if the data is sold off and if so whether it's properly anonymized...


Where do you think the $1-2 comes from? Insurance company is double dipping to optimize those profits!


We have a UHC choice plan at work and it’s been completely invisible, both for me and coworkers I know.

Health insurers are like Congressmen. Most people hate the institution, but are satisfied with their own insurer. https://www.kff.org/affordable-care-act/kff-survey-of-consum...


[flagged]


> I think you're a bad person.

It's not acceptable to comment like this on HN, no matter who it is or what it's in reply to. The guidelines ask us to "assume good faith", "edit out swipes" and "be kind". That applies equally to all of us and in replies to anyone else on HN. If someone posts a comment that's egregious, just flag it, don't abuse the commenter, as it just makes the place worse for everyone.

https://news.ycombinator.com/newsguidelines.html


Comments do not exist in a vacuum. 4chan is a failed experiment

Assuming good faith is a good guideline when there is no evidence of bad faith. If someone usually post racism and corporate bootlicking and is using their collected to Internet points to abuse the good faith of others, the system breaks down.


Please don't sermonize to distract from your own disrespectfulness towards HN and its guidelines.

> usually post racism

If there's evidence of this you should include links in your comment or email us so we can investigate and take action. Rayiner is of South Asian heritage and was born in Asia, which – of course – doesn't preclude racism, but it does at least require us to think twice before presuming his views to come from a place of Western-born white privilege.

> corporate bootlicking

This is an ideologically-charged epithet that’s invoked specifically to be profoundly demeaning. The comment you replied to was simply sharing of his personal experience as a customer of a health insurer. Anyone should be able to do this without being belittled in this way. (Besides that it’s inaccurate; I see Rayiner criticizing corporations often.)

As we've said before, Rayiner takes a lot of arrows on HN because he's about the only well-known participant we have who is notable for regularly espousing conservative/libertarian positions. He also contributes plenty of valuable perspectives from his personal experience as an immigrant, a lawyer, and a compiler-hacker. He's subject to the guidelines like anyone else, and his positions are frequently debated, refuted and flagged by other community members, which is fine. But HN would be the poorer without him.

In this case you're the one who has been escalatory and hostile, and you need to take responsibility for your own conduct before pointing the finger at others.


what kind of mental illness do you have? just wondering


Please don't reply to a bad comment with a worse one, it just drags the whole place down.

https://news.ycombinator.com/newsguidelines.html


I had them via a previous employer and had serious surgery; they asked me to go through the “get a second opinion” process which was getting all the records to their system and then a virtual/video call, but after that, they paid for everything.


I don't. I have their "platinum" insurance, and they still seem to fight any claim we file.

For example, my wife got knee surgery recently, and the doctor recommended we rent a CPM machine to help her knee avoid atrophying. Renting the machine is $200 a week. Insurance said it was "optional" and refused to cover any of it. We ended up buying a used one on eBay for about $900, which is a lot but not insurmountable for us.

It kind of annoys me though, because not all their clients are yuppie software people who have disposable income. A lot of people can't afford to rent a machine for $200 a week or buy one for $900 on eBay, but they do make it much easier for the leg to heal better. Isn't "stuff that most people can't afford but would help with healing" the stated purpose for health insurance? It seems more than a little unfair that my wife's leg is more likely to heal better purely because she's married to a software engineer.

I really have no fucking idea what the difference between the cheap and expensive UHC plans. It sure seems like I'm paying many thousands of dollars more for medical stuff than I was for equivalent services with Anthem. Oh, well, at least my premiums are higher too, so that's fun.

Hopefully obviously I don't advise shooting a CEO for several reasons (both ethical and legal), but I have to say that I was unable to cry many tears when I heard it happened.


>Isn't "stuff that most people can't afford but would help with healing" the stated purpose for health insurance?

I thought the operative term was "medically necessary"? "would help with healing" can theoretically cover everything from protein shakes for knee injuries, to iPads to help with stroke recovery. A CPM machine is on the far end of this, closer to "medically necessary" than the other examples, but you have to draw the line somewhere, so some reasonable-but-theoretically-optional equipment gets excluded.


Sure. I guess I would draw the line in a different place.

There are plenty of things that aren't strictly "necessary" but are still provided by insurance. My wife's painkiller medication isn't strictly necessary, she wouldn't die without it and the leg would probably heal the same way, but they covered that because obviously they should cover that. I feel like a piece of medical equipment like a CPM machine is more necessary than painkillers.


Well first question to my mind is does a CPM actually help? There's a lot of waste in American Healthcare on expensive but fairly useless treatments.

My guess was that a CPM might fall into this category (I did PhD research in bio mechanics in MatSci). So I googled it and it returned a quote:

> Do doctors still use CPM machines? > The machines are no longer widely used because of the multiple studies that found CPM following knee replacement surgery has minimal benefits. However, some surgeons still recommend CPM following knee surgery when the limited pros outweigh the cons in a particular case. (1)

From an insurers perspective it makes sense not to cover a marginally useful piece of equipment. The better use of resources would probably be covering PT where there's movement and weight on the joint.

1: https://www.verywellhealth.com/do-i-need-a-cpm-following-kne...


Fair enough. It was still prescribed by the doctor and I would rather have not paid for it.

Even if its benefits are marginal, they’re probably still more tangible than acupuncture and chiropractic, both of which are apparently covered by my insurance, and the CPM machine probably doesn’t cause a stroke like chiropractic does.


> Isn't "stuff that most people can't afford but would help with healing" the stated purpose for health insurance?

You are confusing "health insurance" with a "system that guarantees healthcare as a human-right". Those are different things.

The purpose of health insurance is:

- To constrain healthcare coverage to the minimum allowed by law or the plan contract, therefore maximizing profit margins.

- To provide a shared risk coverage pool to pay for treatment for catastrophic health events that are unforeseen.

- To provide a product to be used as leverage by employers over employees as part of the "benefits" of a compensation package.

Healthcare as a human-right doesn't exist in the United States unless a health situation has gotten so bad that you end up in the emergency room, which is then legally required to provide you emergency healthcare.


No, but working at a company that was founded by a bunch of former gov employees, the coverage is mostly ok and some having to manage the refusal bs. What was new to me - a gap that went employed -> consulting -> employed - (same company family) - how damn expensive top tier coverage for a family of three really is - $3700/mo. Insane.


if you don't have anything nice to say... :3


[flagged]


> Shooting their CEO caused more people to get claims approved

One, it was a middle manager with a CEO title.

Two, genuine question, can you point out how UnitedHealth’s profits relative to peers changed after the shooting versus before?


I am also not advocating anything but ... wasn't the famous Spock line "needs of the many outweigh the needs of the one". The question is of empirically proving it and that's the challenge. The jury may not be co-opted but the judiciary is. I wonder how do we go about proving this.


Utilitarianism is a dangerous mistress when it comes to justifying moral and ethical transgressions. Sounds great until TPTB decide that the half dozen lives that can be saved with your organs matter more than your one life.


If we followed the rules strictly, and not different rules for the rich, why's that a problem?


You don't see a problem with involuntary organ donations from living people?


Here's a point about all the insurance companies: UHC administers the medical plan on behalf of your employer. For all practical purposes, they are a whipping boy for the real 'man behind the curtain' (your employer).

Your company (for self-funded plans) actually decides what’s covered and what isn’t, sets copays and deductibles, and ultimately saves or spends money on healthcare costs. UHC’s role is to apply those rules, maintain the provider network, and handle the billing and customer service.

If your company offers insurance, there is someone who can tell the "insurance company" to cover the service they are not covering. Usually the HR Benefits Administrator, or 'plan sponsor'. And they do it all the time! If you have a sad story and the budget is ok for the quarter, they will help! If you are a company officer, you can also have whatever your company can afford.


This only applies to large employers. Smaller ones are just presentef a limited list of plans to pick from, and the plans change every year. Most of the time, as a startup, you can’t buy a Mag7 equivalent health plan for any amount of money off the marketplace


It depends. If your employer is part of a self-funded group of other employers, then there is a group of trustees from all the employers that can approve.

If it's a 'fully insured' group plan then the insurance company is technically in charge, but your company can do an Employer-paid exception (aka carve-out reimbursement) to cover something thats getting rejected. They also have the option to purchase add-on policies to add coverage for upper class stuff like fertility treatments, weight loss drugs, or gender-affirming care.


Yeah, I work for a smaller company. I'm not sure which options they omitted but I don't think have the same bargaining power as a BigCo.


Mag7 surely is self insured. They have an amazing risk pool of young people. Probably biggest cost is babies. So in this way employer sponsored health insurance screws the rest of the market, as it "hordes" the best risks. The insurance companies then wail about the cost of the risker pool of those of us stuck in the smaller plans...


There should only be one risk pool which is the whole country. Unfortunately the republicans want to go the other way and push sick people into high risk pools which will be unaffordable for a lot of people


Insurance companies are a whipping boy, but for doctors not your employer.

Doctors charge massively high prices, which is why insurance bills are high. Doctors have the most powerful trade union on the planet and strictly limit residencies, thus limiting new doctor supply and keeping prices super high.


I’m shocked that a company would share how amazingly bad their layer management had become. This may be a great internal blog, but I wouldn’t share it publicly.


On the other hand, I'm impressed that a company is owning up to the problem. Is it a dumb problem to have? Definitely. Are they the only ones to have it? Almost certainly not.

People are going to use the tools at their disposal, and they aren't all going to learn their tools at a high level. Think of every insane misuse of Excel you've ever heard of, for instance.

IT has the choice in this case to mitigate, or limit the access to the tools. Choosing mitigation prevents the growth of shadow IT and helps ensure that IT remains a trusted partner and not an obstacle to be worked around. This reflects well on the company, especially if they then go and provide better training to their users as well.


Yeah: I can usually tell from public information when a company has problems like this, and that makes me disinclined to want to work for them. Seeing how they deal with those problems, though? … Well, in this case, it shows that the company doesn't know how to deal with these problems properly, and thinks ChatGPT is appropriate for write-ups, so I still might not want to work there – but I might bother interviewing there, just to check how deep these problems go. (If they're just a case of "they didn't know better, but they're happy to learn", then I might actually take the job offer: an environment where others are willing to learn without fear of losing face is an environment where I can learn without worrying about that either.)


I'm confused. I had the same initial reaction as you and then read further and it sounds like the image was actually provided by a client?

    > The problematic user image had an astonishing 272 layers, each representing a commit operation.


This sounds like a case of "We are in growth mode and will accept any garbage the customer will throw at us" without calculating the tech debt costs.

As someone who is currently there, it's very frustrating place.


Oh just wait till it’s time for your company to stop the ‘growth mode’ shenanigans and get serious about acceptable levels of tech debt and feature bloat. It’s where we are.

You can’t just flip a switch. There is no “Hey, that was fun, but it’s time to start designing these things with a purpose and vision”. Beyond the totally unreasonable expectations that have been set by Product and C-level- you still have the mountain of tech debt that is coming due and changes slow to a crawl or outages skyrocket or both. Plus, hiring has been based on ‘getting things done’, so you have this group of people who are actually really skilled in hacking things together and getting it out the door. It’s tough and calls for an entire culture shift. How do you stop being a reactionary startup and become vision-based and purposeful organization?


This is the job of tech leadership IMO. People respond to incentive changes. If these items are properly prioritized on the roadmap, and credit and recognition follows tech debt remediation efforts to a similar degree as feature delivery, the work can be done.

But this requires strong tech leadership who can interface well with the C Suite and get buy-in for delaying in feature delivery. In the absence of this buy-in, you pretty much need to control the narrative and create a rogue skunkworks initiative to wrap these improvements _into_ the feature delivery.

Many companies don't have strong tech leadership though, and will perpetually churn VPs and Directors, forever chasing A Change without addressing the culture and incentive system that created that culture.


It's like a car repair company sharing how they dramatically improved ride comfort, speed and fuel usage by using air to fill tyres rather than concrete.


After asking chatgpt for suggestions and trying them all.


Transparency breeds trust.

Sure, it frightens away the short-sighted or particularly excitable people, but anyone who understands how unrealistic perfection is will be comforted by such transparency. Exposing the warts not only sets expectations, but it also assures people that things will (likely) not be just swept under the rug in a company culture of denialism and obfuscation.


From what I understood they provide a kind of shared platform where anyone can run things, and it was one of their clients/users performing the commits.


So they don't set reasonable expectations with the customers and accept any and all garbage. As Ops person, this is a path to Ops hell as customers throw more and more garbage at you and toil dealing with customer problems becomes unbearable.

This is a case of Product Team not working with customers, finding out what is reasonable and allowing system to set reasonable limits.


I would give them some leeway, sometimes you have to learn the hard way. But I was also kind of surprised didn't mention contacting the client anywhere.


I could see a compromise where if there are obscure codecs that may not be as secure, FFmpeg would present a warning before loading the file. This way, the user would have the option to decide whether to load the file or not. By default, potentially malicious files would not be loaded, which could prevent them from being used as part of an exploit. This seems like a reasonable compromise.


> FFmpeg would present a warning

Reminds me of gstreamer plugins being separated in "base", "good", "bad" and "ugly" sets.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: