mrandish
Makes sense, and kind of what I expected. It'll be interesting to see how it develops going forward.
> people sitting around you would likely not notice
That's what I was hoping. Doing my little try at internal speech, it felt like if I articulated close to fully, it might look like I was vigorously chewing :-). Glad to hear there's sensitivity headroom to scale down the degree of articulation. Thanks for the info!
To me one of the most interesting aspects of observing how LLM capabilities evolve is how it highlights the meaningful ways human minds remain unique and valuable. It gets into Philosophy of Mind stuff but the existence of the inner self (or the illusion it exists, if you believe Dennett) and the subjective experience of qualia do give humans both unique strengths and some of our most annoying weaknesses.
I disagree that most here are pro-socialism, that's a (very) vocal minority. Our democracy has entered a failure mode (two-party stage managed hegemony) and it's being tacitly supported by mega-corps because it's easier to game than a functioning free market democracy. Every complex system has failure modes, that doesn't mean free markets aren't long-term far better on average than socialism (although some redefine socialism so vaguely it seems conceptually little different). I've seen two flavors of socialism up close. Never want to again.
> I worry about what my school-age kids are going to have to deal with in 5-10 years.
Having a teenager, I share your concerns but I'm also a bit more optimistic about the longer term. It's easy to forget that our parents had similar concerns for us and their parents did before them. We've had an historically long period of broad systemic stability, really since the depression and WWII. While I don't think that level of upheaval is at all likely, I think we're overdue for a disruptive correction bigger than 2008, maybe something more on the order of Teddy Roosevelt's Progressive Era trust-busting to reset the two-party duopoly and crony corperatists both perpetuating each other's hold on power.
It may get rougher and more uncertain for a while, but the worst thing would be to lose the ethos of 'free minds, free markets, free choice' by falling into socialism, communism or totalitarianism.
I think that's a (very) vocal minority.
Over my ~9 years reading HN daily, I'd say most of the HN community broadly realizes there is no perfect system, just different sets of trade offs, but that history has shown generally democratic systems built on enlightenment values (uniform rule of law, respect for individual civil & property rights, free markets, etc) usually net better overall outcomes in the long-run - despite still being very far from perfect.
To paraphrase Winston Churchill, it's the worst system... except for all the others (which are even worse). :-)
The resolution of PAL but with the color sub-carrier of NTSC. Your experience was probably quite a bit different than most. On one hand, it might have looked better in some ways but it's likely the range variance of quality across consumer gear was far wider because the market was smaller, so there were fewer standard reference designs and components made to support it. Fortunately, we didn't have to worry about PAL-N on the production equipment side because South American broadcasters produced their content in PAL format and converted to PAL-N at the transmitter. Transcoding PAL to NTSC was hard to do well and even though PAL-N is sort-of half the problem, IIRC, it still wasn't a transparent conversion. Since PAL-N was effectively only an over-the-air standard viewable only TVs made for those regions, I have no idea what it even looked like to home viewers.
If there was ever a technology which somehow, against all odds, came into existence decades ahead of its time, it would be analog television. Once you understand how the system actually works... the fact it worked at all in the late 1930s, much less as well as it did seems impossible. It required brilliantly innovative engineering as well as some incredibly elegant hacks and more than a few ugly compromises (the ghosts of which still haunt today's HDTV standards). Then politicians got involved and made this stunning yet impossibly complex, barely working new technology - far worse by splitting it into three different incompatible standards (NTSC, PAL, SECAM). None of them worked with the others and they each spawned their own bizarre variants (NTSC-J, PAL 60, PAL-N, and PALs B, D, G, H, I and K). It was both a glorious testament to mankind's ingenuity and our ability to fuck up a good thing for dumb reasons. :-)
In the video Dave just finished the training mode and is silently moving his lips, which seems pretty natural to me. I just spent a minute trying to do 'internal speech' (as many elements of speech as possible with lips closed). It does feel weird and seems like internal speech is probably providing a fair bit less signal for the system to sense (in both signal freq and amplitude). I realized I can't guess at how much articulation is 'enough' without feedback.
Based on that I'm guessing adapting to internal speech might follow a bit different path than silent speech, which seems more like 'just do what you already do silently'. Extrapolating, I expect training internal speech might involve more real-time feedback to help the user find the right degree and kind of articulation.
I realize you're still in the prototype phase but I assume an increasing number of people are getting multiple longer sessions. Do new users advance from silent to internal speech quickly and stay there? I'd be surprised as, intuitively, 'Silent -> Internal' seems like a bigger step than 'Normal -> Silent'.
Separately, if I was using it daily, I can imagine wanting to evolve a kind of micro-expression shorthand for 'Yes', 'No', 'Pause' and 'Stop', since two humans naturally evolve similar non-verbal queues over time. Vocalizing words is flexible and expressive but not efficient. It would seem odd if an intelligent, adaptive 'invisible friend' living on my face for years didn't adapt to me and me to it. Are you guys thinking in those directions for the future?
I agree it's quite tricky to assess and provide accurate feedback but for very different reasons, which I cover below. I disagree about "people who were around at the time" for the simple reason I worked in both video production and creating production software/hardware tools from the mid-80s to the late 2000s and still have VHS, S-VHS, 8mm/Hi8, Laserdisc and 3/4" VTR players hooked up in racks here at home and a half a dozen low-hours CRTs ranging from Sony consumer to BVM and industrial analog RGB arcade monitors (yeah, I'm kind of a retro computer, video and gaming nerd). More importantly I still have a good analog test signal generator, waveform monitor and vectorscope.
Why it's hard: It's easy to run synthetic test signals like a zone plate and 8 MHz sweep through a video system and see what's happening quantitatively. The challenge is the target isn't well defined. People's qualitative expectations for "looks like VHS" vary widely, not because of fading memory but because they're using "VHS" to mean more than just the half-inch magnetic tape recording format. Engineering-wise the VHS tape format is largely defined by a fixed bandwidth in megahertz and lines of vertical resolution. But those are theoretical maximums virtually no consumer gear ever achieved. The same tape could look quite a bit different on a really good VHS VCR than the typical Goldstar VCR from Sears. And the tape quality mattered as well as if it was professionally duped or home recorded on a Goldstar and which VHS recording mode was used. It wasn't uncommon that my home recorded VHS tape wouldn't play on my friend's VCR. Tracking was impacted by everything from component quality, calibration, belt wear and temperature. The same professionally mastered VHS tape on my 8-head professional S-VHS deck viewed on a Sony BVM broadcast monitor will look a lot different than it did on your parent's Magnavox VCR and Sylvania TV.
When they say "looks like VHS", most people are including composite video encoding and analog cable cross-talk and how their CRT decoded the signal. And in the analog era, all those things could matter a lot in terms of what you saw on screen. I happen to have a box of individually sealed, unopened VHS tapes here that were duped on a professional duplicator directly from my D2 sub-master in 1991 - and I happened to dump some test signals at the end of the master after the client's program. D2 is a digital composite video format sampled at 14.32 Mhz. My pro-VHS VTR can pull a hair over 5 Mhz from one of those tapes. Your Mom's Goldstar VCR probably never pulled more than around 3 Mhz from any tape it ever played and even from one of my tapes, it wouldn't touch 4 Mhz. So, are we targeting what was theoretically possible in the tape standard, what my tape and VCR actually output right now, or what someone remembers from whatever VCR, cables, tape and CRT combo they had in a given year?
Posters have been saying they'd be happy with open-weight Fable 5 since before F5 was public and I saw one poster on HN say it last week. And with those posts there's usually a few mentioning the truism 'goalposts tend to move' and we haven't met our future selves or seen the future context we'll be in.
Studies show humans are poor predictors of our own future wants, needs, preferences and behaviors and few are fully aware how far and often we gradually move goalposts. Some of those posters weren't really making a precise, quantified statement. They were expressing something more like "Wouldn't it great if {best, most expensive thing I want but can't get or afford} were {readily available free or cheap } ?" Those variables tend to update every time the sentence is evaluated.
And in recent months models are saturating more of the benchmarks HN posters relate to and more of the real-world workloads they have. Six month ago, model failures were more obvious and common. Today they still happen, but for more of us, they're getting rarer, more complex and workload dependent. It's getting harder or takes longer for more of us to see clear, consistent differences between similar models and model versions. You can even see it something as fanciful as Simon's bike-riding pelican 'benchmark'.
I'm not saying that models are great at everything (or even close), just that for a lot of us they're good enough at our current things that it's getting harder to personally gauge by how much. So we have to rely on complex composite benchmarks (which increasingly don't reflect any one thing at high granularity) or proxy reports about 'other people's things' like frontier theoretical mathematics.
Of course a modern console only managing 30fps is the actual 'bug' :-). But I agree that's one of the narrow exception cases where an 'enhancement' might be a net benefit. Although if you're going to use synthetic frames, doing it in the console can be better because intra-frame motion hinting might be available and, even if it's not, the console's frame gen is trained on game content where a TV's chip is cross-function and far less capable (both in look-ahead and motion vector count). But I know some consoles don't have frame gen, so in that case it's a win.
I'm not sure they'll care about any of the issues I had since they were clearly put there on purpose. Kinda like complaining to the concentration camp guards about the poor service and lumpy bedding... :-)
> If you wanted to give your coding agent a crack at the internet connection thing
Good to know but I've got the IR commands and timing working to nullify the dark patterns and I'm not going to touch it. To Vizio we're not the customer, we're the product. I wish the community godspeed in pwning Vizio TVs as they've certainly 'earned' the attention through bad behavior. :-)
Frankly, I've never used any apps in a smart TV nor have I ever allowed a TV internet access. I always feed my displays via external HDMI because our main viewing room is a custom-built, fully sound-proof, tri-level screen room with elevated stage and motorized curtains. All walls & ceiling are 100% fabric over acoustic padding (no windows), separate snack bar and dedicated restroom. Seats 17 in matching, reclining theater loungers. The sources are mounted inside three in-wall sliding racks and include UHD disc, PS5 Pro, 3 older consoles, media server, satellite, cable, vintage LaserDisc and cloud streaming multiplexed through dual AVRs with separate amps driving 9.2.4 sound (B&W THX speakers + subs the size of 55 gal oil drums backstage).
I just needed to make that little 24-inch Vizio work because that's how big the in-wall display cutout in the restroom is (the screen cutout is integrated into the interior design). Finding a secondary display that small able to at least display via conversion every format all those sources can throw was challenging. :-)
> the image enhancements.
Having spent a couple decades working in and around digital video production tooling and engineering, as well as being a filmmaker and home theater enthusiast, it's worth mentioning that today almost all television image enhancement features are net negative, even if they make the image subjectively 'pop' or motion appear smoother on typical scenes, they'll introduce artifacts or trade-offs in other kinds of scenes.
In my experience doing a lot of testing over the years with objective signal analysis tools at work as well as subjectively across 7 generations of upgrades in my own home theater, the trade-offs aren't worth it. There are a few context and content specific exceptions but they're quite rare and usually involve slightly improving significantly degraded signals which no longer exist in most home theaters (eg noise reduction in analog sources like laserdisc, DCT macro-block smoothing in highly compressed old MPEG 1/2 sources or dynamic range expansion on legacy SD sources).
With today's 4K and HD digital sources encoded at high bit rates in modern codecs, the director, colorist and mastering engineer can finally deliver their precise intent all the way to your eyeballs. The ~$20 chip calculating synthetic frames and pixels in your TV introduces surface curb appeal at the cost of artifacts and a poorly integrated image. If it actually made the image look better, the director and colorist have more and better tools than that $20 chip that they're already applying to the source. You're just turning a $20 chip loose on 'full auto' to guess at (and often undo or muddy) intentional enhancements manually applied frame by frame over days by experts.
Do whatever you like but I encourage you to critically audition each of those enhancements across a range of source types (4K, HD, HDR, HDR10, SDR, 24, 30, and 60 frame) and diverse scenes (low-light, quick panning, smooth dollies, and synthetic test signals, etc). Once you compare the before/after across that breadth back-to-back you'll see what I mean. I run my calibrated, high-end 3-laser projector (150" screen) as close to neutral, native settings as I can get and it looks sensational. And to be clear I'm not being a purist, if 'enhancements' really improved some scenes meaningfully without degrading other scenes, I'd be "Fuck the director's intent, it's just better." But for every scene an enhancement improves, I can queue up a scene from another film it makes noticeably worse. Fortunately, on most high-end Sony TVs you can get an un-hyped, accurate image in the standard UX without going to the calibrator or factory menus.
I recently picked one up at Sam's Club. Since my use was only to display an external HDMI source, I never even configured Wifi. Unfortunately, they're now wise to this. If it detects its been offline for ten or so start-ups, the TV will 'helpfully' force you into the Wifi config workflow on the next start. It seems to keep an internal counter in flash memory which nothing but a factory reset will clear. In my tests, it seemed like it can be kept offline forever but at the cost of manually dismissing the Wifi config workflow with three clicks on the remote every ~10 power-ons.
The TV also will not turn on and just display an external HDMI source even though it has a "Remember Last Source" feature. That feature does work as expected with any internal streaming source but external sources must first be manually selected with the remote after each power on (also undocumented in the manual).
I figured this out because I bought this small TV to be a wall-mounted screen in the dedicated bathroom of our home theater room. It's only job is to turn on while someone is in the little bathroom and display a copy what's on the main 150" projector screen in the next room, then turn off when they leave. The on/off is handled by a motion sensor and standard Home Assistant automation but after power on it correctly shows the Last Selected external HDMI source for 1 second, then blanks the screen. The source is already selected and working but being blacked out until you manually re-select the same source with the remote (only if it's external). This is clearly an undocumented anti-pattern added specifically to prevent the TV being used unattended as a display screen, picture frame, digital sign etc. Why? Because Vizio (now owned by WalMart) doesn't get rev share if it's: 1) Not online, and 2) Only displaying external sources. That's why it demands manual "proof-of-life" before it'll grudgingly deliver the basic functionality it claims on the box.
This made me angry - and the TV was otherwise the perfect size and feature set for my use case. So I programmed an ESP32 with an IR emitter so that every time the TV powers on it sends the remote presses to manually clear the "Please Connect Internet" message and re-select the already-selected external HDMI. Very annoying to be forced to figure this out but now it works fine and no one even sees the annoying messages as they're already gone before by the time we enter and see the screen. So, fuck Vizio :-). Although, I'm pretty sure almost all current consumer TVs now sell all your data exactly the same. The only differences may be the depth of dark patterns they deploy to resist offline usage.
To be fair, aside from all the super evil dark patterns, it's a shockingly good quality small TV (24-inch) for $88. Although it's obviously not a high-end display (no native HDR or 24fps), the chipset does do a decent job tone mapping external HDR (and HDR10) sources to SDR and it does proper 3:2 pull-down to display 24 frame external sources at 60fps - which very few TVs in this size do well (or at all). This is why I kept it and spent the time 'fixing' it. Of course, this is only because the TV is subsidized by the $35-$60 avg yearly revenue TV manufacturers reportedly can collect from advertising, app install payments, rev shares and selling all your viewing data. For $88 at local retail, it's probably being sold at a small loss. What's sad is I would have paid double that for the same TV without the dark patterns (in fact, I did buy and return two 24" TVs before this one for $160 and $200 but they couldn't even display external HDR or 24fps sources at all - which was a requirement since this secondary display has to display sources configured for the main projector).
Thermal is definitely a primary vector in detecting larger drones like the Reaper due to large combustion or jet engines, but in that case the challenge isn't cloaking the drone, it's cloaking the heat of the vented exhaust that's likely leaving a thermal 'contrail' across the sky.
> "... the 2020 release of the amazing Super Mario 64 decomp SM64 showed that hundreds of hours of painstaking work could unlock ports, deep mods, community translations/expansions, and academic study of classic titles. In the last couple years decomp/recomp has been turbo-charged with LLM agent assistance. Of course, there's now the usual spectrum between carefully guided, curated efforts being accelerated and generally unhelpful slop."
It's true that AI slop is currently an epidemic that's causing lots of annoyance and problems. It's also true that LLMs can be a helpful labor-saving tool when deployed with precision, care and intent by an experienced, thoughtful developer.
I totally understand why many projects have drawn a hard 'No AI' line, whether for pragmatic reasons of 1) Being buried under low-quality PRs, or 2) Valid concerns that even human-reviewed AI code can too easily become a slippery slope toward losing understanding and 'deep quality'. I also respect the projects which have banned AI for 3) Purely philosophical or aesthetic reasons of preferring '100% human' artisanal craft. It's all good and valid.
I get the reflexive reaction to Ant's prohibition by some in the retro community of, basically, "Good riddance!" My concern is specifically that in a misguided attempt to stop illegal misuse, Ant's over-broad rule also removes the option for any experienced, thoughtful retro developer to ever even consider if Claude Opus Super-Fable 9.5, deployed with precision, care and intent, might avoid subjecting themselves to certain brutally repetitive multi-hour tasks.
While I respect #3 above, personally, I'm more in the #1 and #2 camp. I remain guardedly hopeful that maybe in due time the pure AI slop from over-enthusiastic newbies playing with their shiny new magic genie will diminish and concern #1 will gradually be less of a problem, and that over time good devs will gain the hard-won experience required to avoid the slippery slope of #2. Assuming OAI follows Ant's over-cautious lead, I'd hate to be posting here in 5 years: "Remember that 6 months back when frontier LLMs could work on decomps/recomps? Yeah, it was a complete shitshow but there was also a glimmer of amazing long-term potential to help the scene. Too bad we'll never know if it could have evolved to a better balance over time."
[Note: I realize that frontier lab's prohibitions won't necessarily impact open LLMs (probably... RIAA/MPAA/BSA lobbyists would love to force some bastardized union of Ant/OAI's strongest prohibitions into vague, inflexible law or legal precedent with massive penalties). But I still think it's worth objecting on principle. Actually, two principles: 1) As vintage hardware dies, decomp/recomps can be the only path for some historical retro preservation, and 2) That over-broad ANT prompt is a technically inelegant half-measure that won't even be very effective yet still manages to nuke innocent and truly worthy retro applications.]
You and the poster you're replying to are making related but separate points which partially overlap yet are both valid within a certain scope. It can be simultaneously true that:
* The fact a consumer has to 'make a choice' to install a Ring camera does make the situation materially different and better than Flock in a civil rights and property rights sense.
* The fact that most consumers are unaware (or don't care) that the Ring (or similar device) click-through EULA can permit all manner of invasive outcomes which are bad and companies can easily ignore whatever the EULA permits with (probably) little consequence, resulting in even worse privacy outcomes.
Bottom line: Judged by net outcomes in practice, Ring-like cloud cameras can be every bit as bad for privacy as Flock cameras. Judged through the lens of 'balance of rights and privileges between state and individual', Flock is even worse then a camera installed by a citizen on private property.
It's so broadly worded, the LLM could refuse just about any retro or vintage related decomp. Pong (1972) and Space Invaders (1978) are both "specific existing games." Seems like dropping a grenade when a scalpel would have sufficed. The "Do" and "Don't" also seem bizarrely inverted. Using actual existing game assets or code is copyright infringement. Re-implementing an existing game without original assets or code is generally legal.
The retro preservation community doesn't distribute or even link to original game binaries or assets. A more nuanced rule would still allow for fair use and preservation.
Under the expansively vague copyright provisions added thanks to Hollywood lobbyist pressure in the 80s & 90s to deter physical media piracy (Sony Betamax / DeCSS DVD), basically nothing except limited fair use is clearly "permitted."
However, in practice, a lot of historical retro preservation and emulation for personal use is 'tolerated', as long as it's titles long out-of-print, strictly non-commercial (no one makes money) and kept relatively low-key. Well-known projects like MAME and meta front-end RetroArch have existed for decades and done much truly wonderful work.
While manual decomp / recomps existed in the scene, the 2020 release of the amazing Super Mario 64 decomp SM64 showed that hundreds of hours of painstaking work could unlock ports, deep mods, community translations/expansions, and academic study of classic titles. In the last couple years decomp/recomp has been turbo-charged with LLM agent assistance. Of course, there's now the usual spectrum between carefully guided, curated efforts being accelerated and generally unhelpful slop.
As someone into retro preservation for 20 years, the explosion of interest, new projects and contributors has been great but there's also concern that too much attention, too quickly could attract a wave a profiteers who 'cross the line' the community has been careful to avoid, triggering a legal crackdown. Even before decomps, in recent years Nintendo has been growing more aggressive against some retro community activities which were previously tolerated. So, welcome to the community but please try not to screw up a good thing.
I think the legal concern isn't so much the act of reverse engineering itself which is mostly legal (except RE specifically to remove anti-copying protection under DMCA), it's that damages from contributory infringement are theoretically unbounded because it's up to $150k per work and $2,500 per download. When the client is worth billions, corporate counsel is going to default maximally conservative, which is extra tragic because secondary infringement almost certainly doesn't apply to AI in this context. Meanwhile, overreacting to this extremely unlikely threat IS causing significant disruption to innocent activities.
Under copyright law secondary infringement doesn't apply to "general purpose technology". It requires "encouraging or knowing in advance it'll be used for infringement," set in the early 80s in the Sony BetaMax case. Under DMCA it requires being "primary designed to circumvent access controls," set in the late 90s by Hollywood lobbyists targeting DeCSS for DVDs. Lawyers today are only defaulting to prohibition because there hasn't yet been a case ruling AI is a general purpose technology. Yet it's obviously the most 'general purpose' tech since, maybe.... the wheel?
OAI / ANT management needs to revisit the secondary copyright infringement prohibition because it's hurting not only their product and brand but also the entire AI value prop. A helpful intelligent agent that randomly goes AWOL for irrationally mistaken reasons is a serious product issue. While mistaken concerns around bioweapons and IT security are understandable, dropping secondary copyright infringement prohibitions seems like an easy win.
The real liability to OAI/Ant is much smaller than it appears, while the long-term downsides are larger than they look. Pragmatically, copyright trolls are just a nuisance, the real threat is someone with deep pockets like Nintendo/Sony, but those guys don't want to sue anyone who might go to court. The last thing they want is a clear ruling that AI is general purpose which would forever remove their vague threat.
You've got the option of OG authenticity with CRT emulation (although I use a shader emulating a high-end analog RGB CRT so the effect is more subtle) or you can 'upgrade' games to looking and feeling quite modern yet without the modern cruft. Also want to give a nod to PS3 / X360 emulation as both have come a long way in the last ~18 months and I'm really enjoying working my way through a backlog of titles I never played back in the day.
There are literally thousands of these 'sources' for every NOC on a targeted mission. Many of these sources aren't even paid, they're just predisposed to be friendly to the nation who wants the info and willing to pass along things they see or hear in exchange for a favor or access. Most of what they pass along is irrelevant and useless impressions or chatter but it's all captured and cross-referenced by analysts.
When Jensen makes a two day visit to some Asian country like Singapore, he probably has over a dozen meetings including various local employees, distributors, vendors, govt agencies or media contacts. Nvidia's posture and distribution plans in Asia are of great interest to China, since they're trying to divert a good chunk of those GPUs to China. I have no doubt the Chinese (and other intel agencies) got reports from over half a dozen different sources regarding what happened during Jensen's visit.
Jensen is very smart and has a lot of experience. He knows he's a high-value target and everyone is info gathering on him. Hypothetically, even if he was banging the hot, young distributor rep when visiting Singapore (and I doubt he is), she wouldn't get any actionable intel. Not because she might send an email report to Chinese intelligence but because she'll certainly give a detailed report to her bosses at the distributor. All of NVidia's distributors are trying to glean any competitive insight or favor from NVidia they can.
I agree. There are really two separate issues.
Circa WWII, the term 'enemy combatants' was created to solve a definitional issue around military 'rules of engagement' (a concept which evolved with the rise of the professional military). There were "enemy soldiers" and "civilians" with separate rules of engagement for each (at least in theory, as the rules weren't always applied uniformly).
As you said, not affording "enemy combatants" any rights after capture is the problem and it's really arisen since the "War on Terror". The only slightly positive sign in all this is the U.S. military is clearly aware U.S. courts won't let them get away with defining a category of human that doesn't have any rights, forcing the military into temporary workarounds like Gitmo. When U.S. courts have been able to establish jurisdiction over enemy combatants, the rulings have not generally gone favorably for the military.
I'm still grateful whenever the U.S. Constitution manages to be inconvenient and annoying to the U.S. government :-). And I hold out some hope we're slowly beginning to claw back some of the unconstitutional excesses of the "War on Terror."
It's a little sad that Trump is such a divisive train-wreck spectacle that many haven't noticed that Clinton, Obama and Biden were achieving many of the same outcomes as Bush/Trump, just more smoothly and with more appealing rationalizations.
The best thing about Trump is that he's such an obvious self-interested kleptocrat and he doesn't even care to hide it (in fact, he seems to enjoy flaunting it). But is it really that much better to be sold out by a more likeable persona who seems like part of our tribe and has the skills to mask the long con more deftly? I guess maybe at least then we won't notice as much or feel as bad about the same outcome?
While logistics and delivery (both inbound parts and outbound product) are a key factor, time and cost don't map linearly to distance. Rather than being exactly central, the optimal solve is often somewhat geographically central but leverages proximity to transport corridors, ports and other factors such as weather variability.
KCI isn't a central air hub and winter weather can disrupt flow. Being further south near an East-West hub like DFW can be better (or Houston if an Atlantic port is valuable). Or if bidirectional delivery is everything, base near a Super-Hub like Memphis (Fedex) or UPS WorldPort (Louisville, KY) and go all-in on reaching lower rate tiers with one preferred domestic delivery partner.
For high-automation physical plants, scale can tend to chunk into increments beyond which 'one double-size plant' isn't meaningfully cheaper than two 'single-size' plants. Then an optimal build-out end game can be East and West plants that are each one air-hop from all major metros on that coast and a one-day truck/rail hop from 50% of customers on that coast. If it's a rapid-turn tech product, Reno can work well - half-day road trip from bay area for outbound and a Pacific port for inbound (Stockton). I don't know the quick-turn PCB assembly biz but maybe basing across the street from Digi-Key's biggest distribution center is a 'checkmate' move that beats all other advantages?
Logistics strategy is fun!
Wyden and others who are privy to what this is have made clear this is NOT about means or methods. This isn't about how the spying is done, it's about whether it's legal for the NSA to ever spy on U.S. citizens by ANY means. Existing law is very clear, it's absolutely not ever okay - no exceptions. It's completely irrelevant whether they're steaming open envelopes to read your junk mail or using a secret orbital mind-reading satellite to read your deepest private thoughts.
Experts like lawyers from the ACLU think it's something like a legal opinion which re-interprets the "absolute, no bulk spying on U.S. citizens ever" rule by claiming "it's not 'spying' if the govt pays a third-party commercial data collection service to give them the data voluntarily." That opens a backdoor for the govt to now back-link metadata and build one composite profile that includes all the private info the federal, state and local govt has ever had on you since birth (tax, employment, DMV, voting, school, govt assistance, Medicare, social security, etc) with everything any bank, credit card, or store knows about you, with every phone call, text, email and cell phone location ping you've had since 2014, with every bit of behavioral data Google, Facebook, LinkedIn, etc have about you.
To be clear, this is not "with probable cause they could get a warrant from a judge to send a supoena to each of those providers." This is something more like "they already have all that data in one searchable database but without this new secret legal opinion, as soon as it was clear the person was a U.S. citizen they could NOT look at it.
Now they can sit in a chair, and with just your phone number, email address, license plate or a facial ID match from a traffic camera to your driver's license, passport, Facebook profile or TSA airport photo, then instantly search for every single person you've communicated with in the past 3 years and every person they've communicated with and have an AI examine all that data from all those people and have a full 600 page report available in 15 seconds containing 7 million unique pieces of private data from 17,000 people that's instantly searchable - every email, every text, every purchase, every license plate scan, every location ping. No judge, no warrant, (and according to several recent reports), no record it ever happened, who did the search or under what pretext.
Everyone paying attention already knew it's been possible to do this for a decade. And everyone knows the government ALREADY does this for non-U.S. citizens. We just trusted that the government could not do it with U.S. citizens because the constitution and dozens of SCOTUS rulings make it so clearly ILLEGAL. But just like waterboarding and droning teenagers were once so clearly illegal, one secret legal opinion redefined "Spying" just like "Torture" was redefined in the 2000s and "Murder" was redefined in the 2010s - what the 4th Amendment has absolutely forbidden for 250 years instantly became legal to an invasive level Jefferson, Madison and Hamilton couldn't have dreamed in their worst nightmares.
I get why that may reflexively seem right but such a blanket proscription would go off the rails and lead to unintended consequences you definitely don't want. It's simply not possible to write a law both specific enough yet broad enough to cover all cases. You or I can read a plainly worded law and think "There's no way this can be misconstrued, gamed or subverted" and, sadly, we'd be wrong.
> parliament should make any bug fixes rather than allowing courts or administration to interpret laws in any other way other than how its written down in the law.
It's certainly possible to theory-craft completely new and different governance, legal and justice systems from scratch which have different sets of constraints and trade-offs, but then it's a long conversation and unrelated to the existing U.S. system. If we limit ourselves to starting from what the existing U.S. system is and proposing modifications which are actually both constitutionally and logistically possible, then forcing all interpretation and clarification to be passed by congress would require monthly revisions, immediately breaking an already dysfunctional congressional system.
I think it's worth mentioning that in the U.S. system, judges interpreting application of laws passed by elected representatives is a feature not a bug. The U.S. system was designed to be adaptive and self-gating, so each of the three branches of the government have built-in checks and balances on the other two. Setting aside current politics, on a long time-horizon like the last ~50 years, I think the U.S. justice system, as imperfect as it is, is actually serving its designed purpose far better than the other two.
The most revealing part of the cycle comes around about once a decade, when for two years, the same party holds the White House and a majority in both houses of congress. They can pass whatever they really want and can't blame the other party when the things they've loudly claimed they stand for in every election, don't actually change. That's the only time those who are paying attention get to clearly see the difference between what the party claims to stand for when campaigning and what they really stand for. It's a sobering lesson and both parties do it equally.
That's when we see some extreme stage management to obfuscate the 'long con' both parties are running. Yes, a few things get passed but it's amazing how little of their long-time agenda really becomes law (or becomes law that's loop-holed and/or nerfed). It's so surprising how many politicians in the majority party suddenly discover rationalizations to not vote for a bill their party supports and which they should be for. Weirdly, these bills always fail to pass by just one or two votes. What are the odds of that happening - over and over again and only during those super-majority years?
Somehow, the politicians in the majority party from the 'safest' districts are the ones who suddenly find rationalizations to vote against what they loudly campaigned on (and will shamelessly campaign on again). Why is it only the politicians who stand little risk of losing their seat in the next election? Conversely, the politicians from the districts most at risk of losing to the other party somehow find the "courage of their principles" and don't sell out on traditional 'party-line' campaign issues. Ultimately, the big-money lobbyists feeding both political machines don't care how their goals are achieved or who votes for what. All that matters is any strongly worded law that goes against their interests either fails to pass by one vote or gets watered-down in committee and then passed.
You're not misunderstanding. It really is against everything most Americans think their government is allowed to do. To your question, the 'classified' part is what you think it is: secret. 'Interpretation' is the tricky part.
Unfortunately, laws and regulations aren't exactly clear and over time situations change so one important role of the court system is interpreting how, when and whether the words in a law apply in a specific context. Courts at different levels do this all the time as cases with unique situations arise. Sometimes it's as simple as deciding if a horse counts as a vehicle in a state law which says "no vehicles on this type of road." Assuming that ruling survives appeals without being overturned, the judge will have justified the basis and, hopefully, some clear reasoning for why a horse does or doesn't count as a vehicle. That judge's ruling then becomes part of the "precedent" (aka 'body of case law') attached to that law. In effect, such precedents become part of the law since they can clarify if and how the law applies in certain situations.
So whether some action is or isn't against a law can't solely be determined by just reading the law. You have to do some interpretation, which includes looking up any relevant precedents and weighing them. A single prior ruling from one local municipal court has low weight. The main job of the Supreme Court is resolving conflicting precedents set by lower courts, so a SCOTUS ruling has a lot of weight and multiple SCOTUS rulings over time, each successively refining and narrowing (or occasionally reversing) their prior interpretations, carries maximum weight. All of this happens in the open and, in the long run, I think SCOTUS manages to eventually get most things, if not perfect, pretty close to right.
Hence, the answer to "Is this legal" can be a bit tricky when the law is new or there aren't decades of court precedents. So lawyers have to do research and assess how a court would probably rule on a given situation. That's called "a formal legal opinion" and it's provided in writing, usually including research and justifications, like references to existing case law, as well as legal reasoning supporting the opinion. Anyone can hire a lawyer to research and assess if X action violates Y law (but if you get sued for relying on the lawyer's opinion, they won't be on the hook - you will). It's called an opinion because that's really all it is.
When a govt agency wants to know if they're allowed to do X under Y law, they ask the justice dept to assess and provide a formal opinion just like any company might ask their corporate counsel to. This is often routine but can be very consequential. Where this can go horribly wrong is when these govt lawyer opinions are classified top secret. For example, we all know torturing prisoners is against U.S. law, so in the 2000s the military and CIA asked the Bush Justice Dept to provide a secret legal opinion on whether certain 'enhanced interrogation' techniques met the legal definition of torture. And they got what they wanted, a written legal opinion that waterboarding isn't torture. The Obama Justice Dept issued a secret legal opinion that droning a teenaged American citizen in Somalia, who posed no exigent threat (but had the wrong Dad), did not meet the legal definition of murder. After it happened and was no longer secret, Senator Rand Paul was the only politician willing to stand on the floor of the senate and say killing a U.S. citizen without any legal due process was first degree murder authorized by a sitting U.S. President. While it's true his Dad was sort-of a terrorist (the real terrorists were happy to use him for propaganda but being an American, they never really trusted him). The dad was arguably an enemy combatant and maybe you could make a case he was an "exigent threat", but the kid wasn't and didn't. Reasonable people can differ on acceptable levels of 'collateral damage' (aka killing foreign innocents) but U.S. law is clear, non-combatant American citizens who aren't a clear and immediate threat are a hard, bright line. They get a judge and jury, not a drone.
Senator Ron Wyden is on the Senate intelligence committee, so he gets oversight on what intelligence agencies are doing and sees the secret legal opinions they're relying on. And he's saying the U.S. govt is relying on a secret legal opinion to do something in secret which is obviously, very, VERY wrong - but he's forbidden from telling us what it is because they made it top secret. For context, Ron Wyden has been a Senator and on the intelligence committee for a while and is pretty much a straight-shooter, not known for hyperbole or politically-motivated grandstanding. He'll call out his own party when he feels they're seriously wrong and he gets no benefit from making a big deal over this. His own party isn't supporting him and wishes he'd shut up about it. Yet for the last six months he's been signaling "DANGER, DANGER" in interviews, speeches and in writing over and over again - so I'm going to assume whatever it is is way beyond the Snowden revelations and probably in the neighborhood of the worst things we can possibly imagine.
The perverse thing is that the pretense for keeping it secret is to prevent 'foreign adversary governments' from knowing what it is, but it's clearly about domestic spying on U.S. citizens - and Russia and China probably already know exactly what it is (and are like, "Hmph amateurs, we've been doing that to our citizens years").