This is cool. Not a feature request, but I guess my dream version of this in my head since I saw GPT-3 was this + charts, and maybe for Clickhouse.
677 karma · joined February 25, 2013
You can try to do a bottom-up estimate, but how much do you think it costs to moderate the (8th?) largest domain on earth? More or less than that? More or less than what Reddit's current operating income is?
Acting against the interests of their own users will make the platform less valuable in the long-term if moderators and users lose trust in the platform.
This used to be more true than today. So do we collectively perceive US institutions as weaker?
"Let's just move TSMC to the US" is a naive solution to the problem most people with little industry experience arrive at. No real effort is being made to move any significant portion of final fabrication out of Taiwan and into the US, and the latest and greatest processes are being developed exclusively in Taiwan. Progress on 5nm fabs in AZ will probably be excruciatingly slow, dependent on significant numbers of Taiwanese worker exchanges, and produce a chip 2 generations behind cutting edge by the time it's operational.
There's some backlash in Taiwan over "piercing the silicon shield", but it looks like it could end up being a way to get the US and Taiwan to cooperate even more intensely than they already do.
Fun to build, mostly just thinking through the UI, catching up with all the new AI models coming out these days. Would be a ton of work to really polish and monetize an iOS / Android app like this, not interested in taking it much further.
Crazy mix of wholesome and hustler products. Something a lot of people used: https://www.google.com/search?q=site%3Aflurly.com
Not quite high enough fidelity to handle ASL though.
Some issues I ran into testing it, if it's something that interests you:
- Cannot distinguish closed vs. open fingers (always adds gaps between fingers, even if they're touching) (B) - Can't handle crossed fingers (R) - Doesn't seem to like extended vs. curled fingers in some cases (H) - Other failed letters: (Q), (E?), (F?), (G), (Q), (S), (U/V)
But, when signing naturally, it seems to get enough of the shapes and orientation correct enough to understand what I'm seeing. I'm sure there's things it'd trip up on because of some of the above weaknesses in detecting hand shapes but it does seem to get movement, orientation, and position "good enough".
If you ask GPT-3 to correct an error it always confidently tells you you're wrong and it made no mistake (in my experience).
It's funny but I imagine more annoying when it's eventually integrated into a paid service people rely on.
Part of the story is that we had a massive confluence of three deflationary pressures through the 90s to present: Labor deflation, Tech deflation, and Resource deflation.
Our biggest bulge in working-age westerners (Boomers) hit their peak skill years in massive numbers, keeping wages for skilled labor low & easy to source. The collapse of the Soviet Union dumped massive quantities of underpriced commodities and mineral resources on western markets, as their high-end manufacturing collapsed. And, shifting manufacturing to China & the far-east kept wages low for decades (on the low end). All of this, combined with the deflationary pressure of technology & automation (increasing worker productivity), brought prices for most goods down dramatically.
But, all of these trends are reversing in the long-term. Wouldn't call it a systems collapse though. Will probably disproportionately affect those on the margins.
This should create a lot of demand for even more systems which increase worker productivity– AI, automation, etc. -- which should be great for all of us in the tech industry, because there aren't a lot of other solutions to the problem (unless you can figure out how to massively lower demand while people are earning more than ever due to skilled labor shortages).
Even with the wage increases we saw in the last couple of years, a majority of these wage gains are probably ahead of us (potentially delayed by any cyclical activity).
It's probably nothing for you to worry about. But, it's likely to be a long-term trend that becomes the new normal for decades to come, and this will shape how we live for a few decades, at least.
https://markets.businessinsider.com/news/commodities/opec-mi...
The Linux desktop seems to tweak some things in the UI to be friendlier toward users coming from Windows. Not sure if that's a valve thing or not, can't say I'm a distro guru.
It's not super performant for me on a 4k display, docked. But, it should all work with a nice usb-c monitor, keyboard and mouse, and a dock off Amazon.
Guaranteed there will be something funky in the distro that gets in your way at some point though.
Oh– It's a buggy video call solution and poor replacement for Slack? Yeah, not important.
They're at over 270m users in an ecosystem of 350m Office 365 users. There's no "less buggy version of Teams" out there with market share to attract by making minor improvements to the software.
It does exactly what it's supposed to, and what Skype for Business already did. The biggest thing moving the roadmap forward is Teams capability to do "Office 365 Enablement" - i.e. win new users into the Office 365 ecosystem.
There's a lot of things Microsoft can do to accomplish that. Some of them may be new Teams features. Some of them are tweaks to pricing and bundling and other unsexy things like that. Some of them may eventually be bug fixes or fun features that make Teams more enjoyable to use.
But, at this point, your best chance of getting minor features built, is to hope internal Microsoft employees find it annoying enough to prioritize, just so they can see the feature fixed.
Because Teams does what it's supposed to, decently well, and well enough to succeed massively in the marketplace for "generic video call software I don't have to pay a bunch of extra money for".
The right design would have the ability to reply to another device with an advertising beacon, which could work in a limited set of circumstances, with some tweaks to how the AirTag protocol currently works. Like, if you stuck your wallet in your pocket next to your phone or on a table next to your laptop.
RF ambient energy harvesters also exist, which use more generalized RF to gather energy in a generic way. Where RFID sends the signal intentionally that powers the RFID reply beacon, you could theoretically find something in the ISM that gives you enough power to somewhat reliably harvest that instead.
https://www.digikey.com/en/articles/rf-energy-harvesting-bat...
Not sure about the advertising beacon interval but if it's flexible enough you might be able to join the two concepts. The limitation is likely figuring out the right ISM band, and the right package (coil size) to harvest enough energy reliably to send out your advertising beacons.
Distance and directionality (phase) once again affects energy output, but also, you would need to build a coil to a specific size to the frequency you're trying to harvest. Not sure these ever match up for this application. Higher frequencies would allow for smaller coils, but with lower power output efficiency.
The math behind this is wrapped up in the definition of the Q factor, and this is very similar to how you understand radio propagation in general, if you're looking for something to research further.
But, my ELI5 to myself: BLE beacons are already using ISM, so if your goal is to retransmit an ISM signal that beats the noise floor you're using to harvest RF in the same range, efficiency is going to have to be phenomenal in your system design (or take advantage of some power differential).
There's been some interesting projects and papers on this in the last 10 years or so, and the UW had an interesting retransmission concept a year or two ago they built. Links appear to be broken now though.
Why? It's the cheapest product w/ a full Mac experience. You don't want to plug in a mouse and windows keyboard w/ a random Monitor and save $200, if you've never used a Mac before. You're not going to experience using a MacBook w/ gestures and all the bells and whistles a normal user would have, and that might end up holding you back.
It's like not having a real iOS / Android test device & testing in emulators – you can verify the buttons do what you expect, but you'll never discover the app "feels broken" without Android back button support.
I'm assuming you have enough traction or a serious reason to pursue working on this app, and you're looking to support it long-term, not just a quick port for a project or contract work. If this is starting to sound like too much work for your project, obviously my experience doesn't apply. MacInCloud is a good service if this is like, a contract gig or something minor, or you're not ready to buy an expensive test device.
At the price point though, it's hard to find a fast, Retina Intel model that saves you enough money to be worth it.
There's also some minor packaging quirks for an Electron "universal" installer that will work on both an Intel and M1 Mac. If you get it to build for universal it'll also run on an Intel Mac.
If it feels like a solid app on the 2020 M1 Air, it'll feel right at home on the latest unannounced $4,000 MacBook Pro coming out next month as well. You just have to imagine the notch.
USB-C charging too, so you won't have to worry about replacing a proprietary charger someday, or buying accessories that won't work on your other devices.
Here's some common OS differences to look out for, as well, during development:
- Test your app full-screen and in windowed mode. There's some interaction quirks you might not be used to when you go full-screen and end up with multiple workspaces
- Understand the drag and drop app installation experience, if you've never seen that before. Counterintuitively, most people probably won't prefer to get your app from the macOS App Store (but you probably want to avoid an unsigned app from an "unidentified developer" – could impact your conversion rates)
- If you use menubar functionality test a few other apps to get a feel for the norms around this – people seem to hate electron-looking apps that look non-native in the menubar. Might be a nice thing to add, if it improves your app experience (seems like a mostly unused feature on Windows)