Had been a happy Claude Pro subscriber since March 2025, and a happy Claude Max subscriber since May 2025. Cancelled my sub, will be deleting my account. This is disgusting coming from the same org that pretended to push back on the Pentagon over mass surveillance concerns. Complete hypocrites. Nemotron 3 Super and Qwen 3.5 122B A10B running on my local hardware are more than sufficient for my needs. I'll miss newer Opus models - probably will end up trying them out a little via third party API platforms that don't abuse their users like this, but seriously, fuck Anthropic.
Happy IBKR customer here. ForecastTrader has absolutely horrific liquidity outside of maybe 30-40 large contracts. The rest is all market makers that only offer 10-100 or so shares at each price point before bumping up a penny or two. No knock on IBKR as a whole, but you can't even effectively buy on most events or outcomes without slippage eating away your entire edge, and forget about real serious positions above a few grand entirely outside of those 30-40 big contracts.
Bitcoin exists. Completely permissionless, anyone on earth can use it. Easier to accept as a merchant than any third party integration. Doesn't require you to trust any government at all.
Cool, but unfortunately, it has the same same drawbacks as cash. If you get scammed, accidentally pay too much or lose your wallet you will never get it back. I sleep safer knowing that there is some protection in the banking system against losing money all of sudden.
Unfortunately it's also pretty clunky for tax reasons in many places and inherently deflationary (and as such problematic from an economic point of view).
Sure, great if you don't trust your government or whoever issues your local currency, but if you can, there are better alternatives. Trust is an asset, not just a liability.
It might not always be warranted, but where it was, increased trust in society, institutions, and systems has been the enabler for economic growth and human development in the past centuries. Talk it down at your own (or more accurately, at all our) peril.
Economic growth and human development over the last several centuries has been the result of a complex web of interleaved prerequisites, that said, trust wasn't one them.
People trusted institutions for thousands of years prior to the scientific revolution. Europe had plenty of trust in religious institutions between the collapse of the Roman empire and the scientific revolution, and you know what it got them? Superstition, witch hunts, barbarism in the name of proselytizing, failed pandemic responses, and a near complete stall in technological and scientific breakthroughs for a millennium.
What the scientific revolution brought us was the decision to not trust, but to reason, to measure, to hypothesize, to verify. Facts matter. Humans are stupid and it is human nature to place trust exactly where trust is least warranted.
"Economic growth and human development over the last several centuries has been the result of"
Fossil fuels...most of the growth from 1800-1970 was due to fossil fuels. Not sure why this is such a mystery to so many. Makes sense when you think about it from a physics POV. You use energy to move things, to make things, to travel to buy things, etc. Heck, the middle class wasn't a concept until the industrial revolution which was caused by...say it with me...fossil fuels.
Like I said, complex web of interleaved prerequisites. Without the scientific revolution, hydrocarbons would remain almost entirely untapped.
But yes, energy was absolutely one of those prerequisites. Fun fact (you're probably already aware, but for other readers): there is a strong positive correlation between national energy consumption and national economic output.
In addition to this conversation already having been started at https://news.ycombinator.com/item?id=47735348 yesterday, MiniMax M2.7 is not open source. The open weights have been released, which is definitely good and follows some of the spirit of open source, but isn't the same thing.
While an open-source model is obviously preferable to an open-weights model, the difference between the two is much less important than the difference between an open-weights model and a proprietary model.
There are much more people who are interested only in doing model inference, for which an open-weights model is sufficient to avoid the uncertainties and costs associated with a subscription, and for enabling them to make and use better model harnesses than those offered commercially (better by being more suitable for their specific needs), than people who also want to do model training, for which an open-source model would be needed.
Absolutely - I'm one of these types of people who just want local inference myself. I have a Strix Halo rig and I'm thrilled to have Minimax M2.7 weights to run locally. Like I said, this is still an unambiguously good thing, and follows some of the spirit of open source.
Just know that Minimax M2.7 is offered with a noncommercial license. If you use it for commercial purposes, you may be on the hook, liability-wise.
>around 60% of rich-world respondents say they would be willing to give up 0.5% of their income if that were enough to end extreme poverty.
If they really were, they already be doing it, and it would be a solved issue. For many folks, it's a lot easier to say 'yes' to a survey about whether you would give your own money to the poor than it is to actually give your own money the poor.
When it happens collectively e.g. through taxes, you get somewhat of a middle ground between the survey and giving your own money (directly and visibly).
There seems to be some kind of international target of 0.7% of GNI (~GDP) for developmental aid already, which governments often don't meet fully but come close to (e.g. https://commonslibrary.parliament.uk/research-briefings/sn03...), so the 0.5% would probably be viable in tax form.
>I don't think Microsoft cares (about anything else than making money), but there are plenty of (state) actors that can influence the decision-making at Microsoft when it comes to these issues.
Microsoft the corporation may only care about making money, but a lot of very high ranking folks within MS Security aren't just friendly to intelligence agencies, they take genuine pride in helping intelligence agencies. They're the kinds of people who saw nothing wrong or objectionable with PRISM whatsoever, they were just mad they got caught, and that the end user (who they believe had no right to even know about it) found out anyway. The kind of people who openly defend the legitimacy of the FISA court.
This aren't baseless accusations, this comes from first-hand experience interacting with and talking to several of them. Charlie Bell literally kept a CIA mug on a shelf behind him, prominently visible during Teams calls, as if to brag.
Remember - Microsoft was the very first company on the NSA's own internal slide deck depicting a timeline of PRISM collection capabilities by platform, started all the way back in 2007. All companies on that slide may have been compelled to assist with national security letters. Some were just more eager than others to betray the privacy and trust of their own customers and end-users.
I was always convinced that Skype was bought by microsoft so CIA/US intelligence agencies to have listening capabilities.
The first thing Microsoft did after the Skype purchase was making it easier to tap into the calls by removing p2p calling and routing calls using centralized servers.
That's my experience with most computer security folks as well, and tech companies who sell security products. Cloak-and-dagger stuff running 24x7 in their heads.
There are quite a few extremely talented security folks who are more or less the polar opposite, who view people like Edward Snowden and Julian Assange as heroes, the NSA as guilty of treason, as James Clapper as guilty of perjury, even inside of corporations like Microsoft.
The catch is, views like those must be kept to a fairly modest level by the people who hold them. Discussing them with ideologically aligned colleagues may be fine, but for example, when someone makes statements or asks questions with such pro-privacy framing on stage directly to security leadership at internal company conferences, that is a quick way to a severance package not only for the person on stage, but also for dozens of folks in the audience who clapped a little too enthusiastically at the onstage remarks.
Because H1b is an arrangement that more or less amounts to indentured servitude where vulnerable people have their visa status glued to their at-will employment agreement, resulting in a dynamic where employers can and frequently do expect unpaid overtime, fewer sick days, and otherwise disproportionately greater value from h1b employees, and those who fail to meet these unfair expectations are let go and effectively evicted from the country as it is extraordinarily rare to to secure another h1b job within 60 days.
The number on two paystubs can be the exact same while one person is being brutally overworked and the other given a leisurely, comfortable WLB, which effectively amounts to underpaying the foreign labor, per unit of output, devaluing each unit of labor of domestic output.
H1b is tied to employment, not to the employer. You can change employers on the same H1.
It’s not great. But this is similar to how health insurance is tied to employment, not to the employer. Both citizens and H1 employees experience the same abuse here
Adding two extra bits to each octet, making each octet range from a still memorable 0-1023 rather than 0-255, would result in an addressing scheme 256x larger than all of IPv4 combined. The entire internet works fine even when IPv4 was nominally exhausted. NAT and CGNAT are not sins, they're not crimes, and there's no rational reason to be as disgusted with them as IPv6 fans are. Even then, IPv4 exhaustion wasn't really a true technical problem in the first place, it was an allocation problem. There are huge /8 blocks of public IPv4 space that remain almost entirely unused to this day.
The reason I'm an IPv4 advocate in the IPv4/IPv6 war is that the problem was "we're out of address", not "your thermostat should be natively routable from every single smartphone on the planet by default and inbound firewalls should become everyone's responsibility to configure for every device they own".
CGNAT is a feature, not a bug. Blending in with the crowd with a dynamic WAN IP is a helpful boost to privacy, even if not a one-stop solution. IPv6 giving everyone a globally unique, stable address by default is a regression in everyone's default privacy, and effectively a death sentence for the privacy of non-technical users who aren't capable of configuring privacy extensions. It's a wet dream for shady data brokers, intelligence agencies, organized crime, and script kiddies alike - all adversaries / attackers in threat modelling scenarios.
IPv6 adds configuration surface I don't want. Privacy extensions, temporary addresses, RA flags, NDP, DHCPv6 vs SLAAC — these are problems I don't have with IPv4. More features means more opportunities to footgun with misconfigurations, being forced to waste my time learning and understanding the nuances of each (in again, what amounts to system I want nothing to do with).
"Reaching your own stuff" is already a solved problem, too. Tailscale/Headscale gives you authenticated, encrypted, NAT-traversing connectivity. It's better than being globally routable. It's also opt-in for anyone who wants it, and not forced on anyone, unlike the IPv6 transition.
I don't have your problems with ipv6, and I'm actively using it.
I don't have to rely on extra commercial entities to be able to reach my network.
I did have a problem with hosting my own shit because my ISP by default does cgnat. That cost me an hour of my life to convince a party to give what used to be normal, end to end connectivity.
yes you do, the control plane is closed. Only reverse engineered by the headscale project. The control plane is necessary for the peers to find each other. If you need to rely on such a crucial part being reverse-engineered, than yes, I think it's fair to say you are ultimately relying on commercial entities.
Highly disagree. Middleboxes are a huge problem on global scale and have frozen any innovation below application layer. TCP and UDP even that they are on software not hardware layer cannot be updated or changed, see MPTCP efforts or QUIC giving up and building on top of UDP.
If this is so much privacy problem, IPv6 is there for many years reaching 50%+ deployments in some countries, I bet there should be concrete examples of such breaches and papers written.
> Reaching your own stuff is already a solved problem, too. Tailscale/Headscale
No address to receive communication - no problem install an app that would proxy it through someone who has the address.
Tailscale/Headscale is great, using it daily, but they are not solution to the huge already build global network created to connect devices not connecting devices because lack of digits. Global is key here.
Wouldn't easy and accessible self-hosting be a major privacy win if that's your primary concern? Sounds much more private to run a Minecraft and Mumble server on an old laptop in a friend group than paying a commercial entity like a hosting provider to know about it and have a back door.
Easy and accessible self hosting isn't the primary concern.
It's much more private and secure to run that Minecraft or Mumble server on an encrypted overlay network like via headscale + tailscale rather than exposing both services directly to the entire planet.
But again, the primary concern was only ever address space.
What I tried to express was privacy being the primary concern. The easy and accessible self-hosting on old hardware would be the uses of a home network beyond superficialities like consumption and commerce. Privacy wise headscale as a solution is still not quite there, because it either necessitates an additional third party to host the headscale server and know about all my friends, or jank like dynDNS.
The additional security gained by getting everyone involved to set up and configure separate VPNs for different community utilities is not worth it.
I wouldn't call port forwarding "huge problems". It's only one minor router setting and if you don't want to deal with it, there's the abomination called upnp.
> I wouldn't call port forwarding "huge problems".
Port forwarding has massive problems if you're running applications expecting certain ports and need multiple hosts to have public access to those ports.
I can't think of any application that needs "certain ports" and it's not a public server application. Why would a home net have a public server app? And why would it have more than one?
Public servers would use a firewall, a load balancer, possibly a reverse proxy that does TLS too, between the internet and real app servers anyway. If anyone needs to host multiple servers from their home net, I would expect them to know this.
Maybe I've got a media server with a web interface, my owncloud instance, a web UI to my SDRs, and some other web app I'd like to have public. Now I need a reverse proxy as well, yet another piece of infrastructure to maintain. Trying to firewall it so only my friend's house can use the media server and I can connect to the owncloud from the office and thenSDR from anywhere is messier and more complicated. Or I can just have each device have its own public firewall rules and not have to maintain a reverse proxy as well and deal with its own complications.
I've had game consoles with matchmaking issues with multiple consoles fighting over the same collection of ports
I can't say anything about consoles. I never had one, but I accept what you say is true.
About the other stuff... well... maybe I'm too old, but I think you're doing it wrong.
1) Those aren't public services and they are NOT expecting certain ports. Public means that a random guy on the internet is able to connect to your service without any prior knowledge about it. HTTP(S) for web pages is one of those services. Owncloud is not, even if it uses HTTPS.
2) Since they're not public, each of those services can have its own port.
3) You shouldn't expose that stuff on the internet, even with a firewall. Use a VPN if you want access when you're away. Or a SSH tunnel. Or at least port knocking.
They're services I wish to consume outside my home
> they are NOT expecting certain ports
Damn near everything expects to be HTTPS/443
> Public means that a random guy on the internet is able to connect to your service without any prior knowledge about it
Public just means I have the option to allowing that traffic or not. I can choose to filter it through a lot of different means. IP filtering, authentication, etc.
> each of those services can have its own port
So I change having easy to remember names for weird odd ports and hope all the other applications handle these changes fine and deal with odd port forwarding issues.
> You shouldn't expose that stuff on the internet, even with a firewall. Use a VPN if you want access when you're away. Or a SSH tunnel. Or at least port knocking.
Wouldn't it be nice to have the option to not have to rely on such things and just be able to connect to things directly? And to have that identity stay consistent both in and out if the "local" network?
> They're services I wish to consume outside my home
Like I said, not public.
> Damn near everything expects to be HTTPS/443
So let it. Forward your own chosen port to 443 of the machine with the service.
> Public just means I have the option to allowing that traffic or not.
Nope. That's not what it means.
> So I change having easy to remember names for weird odd ports [...]
Easy to remember names comes from some DNS service, which can't work unless your public IP address stays the same and you can run your own public NS and can receive recursion from your main NS provider. In most parts of the world, I think, IP changes every time you restart the router, even IPv6. So what you have there isn't exactly typical for home users. That one extra service and complication that you don't actually need.
I don't have to remember anything HTTP(S). No ports, no IPs, no names, nothing. I use bookmarks.
> Wouldn't it be nice to have the option to not have to rely on such things and just be able to connect to things directly?
Wouldn't it be nice if there were no bad guys on the internet, no bots, and no zero days?
> ipv6 is awesome.
For Google and for you, maybe. But let me know how much you like it after your first cryptolocker. I suppose your backup storage is also public on the internet?
Outside the home, in other words in networks other than at home. Potentially without VPNs. Accessible from other public IP addresses, potentially limited scopes of those.
It seems you're thinking that allowing the traffic from other public networks is an all or nothing thing. That either you allow all public network traffic or none of it. That's just not true. If I know my office network is one prefix, and I know my friend's house is another prefix, and I know my cellular carrier in my city is usually this prefix, I can greatly limit the scope of access. It doesn't have to be an all or nothing, either its open to every single other device or only local devices, I can define exactly which networks or devices I want to allow the traffic from or not.
If I wanted to just let my friend's network audio receiver to connect to my music server, I can add it's public ip address and allow that traffic. No VPNs, no tunnels, no proxies, no non-standard port assignments, just directly allowing it to talk to the music server. All through the public internet, but still locking down my music server to just local traffic and his remote network audio receiver.
That's all still "public" networking.
> In most parts of the world, I think, IP changes every time you restart the router, even IPv6.
Often not with IPv6, prefixes stay pretty consistent usually. Not always true, but often true. I've had the same prefix for many, many years at multiple locations and multiple providers without having any kind of payment for static IP addresses.
> So let it. Forward your own chosen port to 443 of the machine with the service.
So now I have to remember the port for local and the port for remote along with different IP addresses for both, or I can just use the normal service port and the same IP address either way, and have one DNS entry for that IP address and it works anywhere I want it to.
> But let me know how much you like it after your first cryptolocker.
Entirely a tangential, unrelated point once you understand how things like "firewalls" actually work.
> I suppose your backup storage is also public on the internet
My most important backups are offline and offsite. But for other stuff, yes, it has a publicly routable IP address. Its not generally accessible publicly though. There are these things called "firewalls", they're really quite neat. One should also think about authn/authz as well next time you're working on your storage solutions as well, it'll do a lot to prevent cryptolocker issues you're so worried about.
For instance, AWS S3 is all technically accessible publicly. Its locked down by policies, not by NAT limitations. And yet its generally seen as a very secure place to store things, assuming one has the right policies in place. It doesn't take it being behind a NAT to be secure, because if that's the major part of your security posture preventing your stuff from getting cryptolocker'd you're doing things very, very wrong.
> I use bookmarks
Ok, and you hop on someone else's computer and...where are my bookmarks? Oops! Or I want to connect back to my media server from a friend's streaming device...where are my bookmarks? Oops!
And once again I bring up things like game consoles and other P2P applications which just work far better with actual publicly routable IP addresses directly. Strict NAT configurations will often cripple these services and good luck trying to have multiple consoles operating at once. CGNAT makes online gaming for some of these consoles just completely unusable, but if we just supported IPv6 it would have no problems.
> If I know my office network is one prefix, and I know my friend's house is another prefix, and I know my cellular carrier in my city is usually this prefix, I can greatly limit the scope of access.
But it's still private use, not a public service. The fact that you access it from the internet side doesn't make it public.
What you are describing here is in stark contradiction with your claimed easy to use IPv6. Why go through all this trouble if you wanted something easy to use as you describe IPv6?
> So now I have to remember the port for local and the port for remote along with different IP addresses for both, or I can just use the normal service port and the same IP address either way, and have one DNS entry for that IP address and it works anywhere I want it to.
No, you remember just your global DNS name which would be independent of ISP, geo location or router used, and work with dynamic allocated IPs too, and remember or bookmark a port number for each service. Reflection in the router would make that combination (name+port) work from inside the LAN too.
>> But let me know how much you like it after your first cryptolocker.
> Entirely a tangential, unrelated point once you understand how things like "firewalls" actually work.
Oh, I know how they work, at least in Linux, I do. It's a 1 (one!) wrong setting away from exposing everything using IPv6 inside the LAN.
> For instance, AWS S3 [...] And yet its generally seen as a very secure place to store things
Do you really belive that? Really?
> Or I want to connect back to my media server from a friend's streaming device...where are my bookmarks?
OMG! And let his favorite cloud providers index my music collection?? No thank you! I ask for the 3.5 jack cable and connect my phone. Or BT pairing code.
> The fact that you access it from the internet side doesn't make it public
A service I'm using its public IP for, routing through public network connections over the public internet, but somehow its not public networking its private networking despite private networks not really being involved. Got it. Having firewall rules suddenly makes it private networking, somehow.
> remember or bookmark a port number for each service
Or just don't, because I've got quintillions of public IP addresses just lying around. I can even have multiple instances of the same service running on the same box all running the same standard port numbers because I can just grab yet another IP address all day long. Why limit myself to having to memorize weird ports when I can just use the standard ones?
> A service I'm using its public IP for, routing through public network connections over the public internet, but somehow its not public networking its private networking despite private networks not really being involved. Got it. Having firewall rules suddenly makes it private networking, somehow.
It's a private service exposed to public networks when it shouldn't be. That's not how it's done. You are taking risks. I'm sure others would agree with me if this article wasn't this old.
> Why limit myself to having to memorize weird ports when I can just use the standard ones?
For the reasons I explained before, mainly LAN security and DNS not working with dynamic IP allocations. But, go ahead, have it your way. Each of us have our own priorities in life. Convenience is a valid choice. Keep those offline backups updated.
Its not how it was done because of NAT. You just couldn't have done it that way for a long time. We don't need NAT anymore.
> For the reasons I explained before, mainly LAN security and DNS not working with dynamic IP allocations
But once again, its not actually changing the security characteristics at all compared to choosing weird ports and having to deal with reflection issues. The service is still opened either way, just one requires you to choose other ports.
reply