The pedantic correction is important in this case: "cancellation" is a private action between citizens, this is "censorship", which is done at the behest of the government. The former can be arguably but reasonably understood as a market finding a balance between two opposing arguments, both of which have a first amendment right (i.e. I don't have to repeat others' words if I don't want to, even if I'm doing it out of self interest).
The government has no such right. Congress shall make no law abridging the freedom of speech.
You didn't update your card after Colbert? Of course Jimmy was next to go. Just look at the comments from Trump directly at Kimmel. Nothing happened after Colbert which just emboldened for this move. This move will also go unchallenged which makes me think the next two shows will be right around the corner.
They say during the gold rush, the people selling shovels made more money than the miners themselves. AI has a similar pattern. Those profiting from AI hype it relentlessly. Meanwhile the butthurts with nothing to sell become the loudest critics, just to stay relevant.
We get it guys. AI sucks and you don't like it. You need not turn yourself into a parrot. Nobody's in the market for your outrage.
I think Signal is the safest choice. If you want to be absolutely sure, host your own service, and hope you know how to make it have airtight security.
This is actually probably the best thing OpenWRT have done in quite a while. I got two (one for a backup) and I've been super happy. I've happily used TP-Link, GL.iNet, and Raspberry Pi 4 devices, with OpenWRT in the past, but nothing beats the "it just works" aspect of a first party device designed for this, fully open, and a reasonable price.
Most off-the-shelf devices have too slow of CPU for a low latency/buffer router. The Raspberry Pi 4 is easily fast enough but needs to use USB3 network adapters which require packages not in the default rpi4 OpenWRT image. Not insurmountable, but a consistent pain every upgrade.
I ended up building my own image of OpenWRT to make the package hell better on upgrade, and as a bonus, I was able to build in my configs too. Recovering from a failed Pi was as simple as flashing the most-recently-built image. Upgrading just required rebuilding the image (and resolving whatever went wrong, of course, though it was usually pretty light). As a bonus, it's easy to swap SD cards on the Pi so I can have the last "known good" config available while taking the update.
Now I run OpenWRT on one of those x86 mini PC boxes with 4x 2.5GBe Intel NICs because my wirespeed is 2 Gbps symmetric, so I needed just a bit more oomph than the Pi could provide. The hardware is somehow even _less_ reliable than a Pi 4 - I'm already on my third machine in 3 years. I would love to find something more reliable.
OpenWRT has also made it incredibly easy to package in any arbitrary pkg into image downloads from their website. You don't need your own build infrastructure now.
I'm curious what your experience would be like with a Pi5/CM5 solution using PCIe for your ethernet. It is pretty easy to have spare boards and SD cards around for Pi setups. I've had good reliability with Pi setups using good passive cooling (no fan to die).
Not a huge fan of the design decisions on that one. $250 target makes it a hard sell to anyone but network nerds. At $100, I would have no issues making that the default recommendation for anyone, regardless of technical knowledge. Being a premium point requires justifications beyond open source warm and fuzzies.
Network enthusiasts are likely to already have separate switches and WiFi points. Let the router just route.
The CPU in such switches is far too slow to do any data plane operations, so performance is entirely due to the hardware switching. Replacing Ubiquiti's OS with OpenWRT just gets you a different management interface for configuring the hardware features that actually handle traffic. (Unless for some reason you desperately want to have a VPN endpoint that would be limited to a few Mb/s at best.)
Ubiquiti make great hardware, however a lot of people hate the software as it is buggy and inconsistent. This at least gives those people options to run something a bit more opoen and manageable under the hood.
I’m struggling to see a subjective version of this that is ethical?
Profit maximizing sure but that’s not ethical if you’re knowingly harming others. So I guess you’re helping your shareholders which is the ethical thing to do since the benefit to them outweighs the harm to the kids?
I’m just saying that some companies might release more information if the reaction wasn’t always adversarial. It’s not just meta. There’s a constant demand for outrage against big companies.
I don't want to beat a dead horse, since sibling commenters have covered this, but I'd implore you to imagine the spectrum of reactions which Meta _could_ have had when discovering their research indicated they were having a negative impact on people.
Some of those reactions on that spectrum would lead to greater human flourishing and well-being, others of those reactions would lead to the opposite. Now think about the reaction they actually _did_ have. Where on the aforementioned spectrum would their actual reaction fall?
Zooming out, how have they reacted to similar circumstances in the past when their own internal research or data indicated a negative impact on people?
The continued "outrage" is that they've exhibited a recurrent pattern across myriad occurrences.
