🧵 View Thread
🧵 Thread (104 tweets)
"According to a Kinsa Insights case study, Clorox, a household name for disinfectant, used Kinsa zip code level targeting data to inform an advertising strategy."https://t.co/4ZsTasoG6MGross.
"the way user data is actually pretty poorly protected and able to be hacked easily, the fact that nowhere do they tell users who shell out $ for the hardware product that their health data will be used to sell targeted ads to them, etc."https://t.co/sTlk5EL7c9🤨
Gotta love a health company that tracks your health to sell to advertisers. One where the VP of Data science specializes in "multi-armed bandit applications in marketing" https://t.co/lY39If7uzeSketchhttps://t.co/xmoJlVuxVU
I just want to not be tracked by a massive distributed data-mining ecosystem hell-bent on predicting what I'm going to do to sell me ads.https://t.co/FTB2Hp0MfZ
I fear I may not have enough lantern oil.https://t.co/7rEwd82ghY
A company, known for using their health tracking data to target advertisements, partners with leading health brands to get health tracking tools directly into your school!https://t.co/AnbDTpXsZd
My, grandma, what big eyes you have.All the better to watch you with, my dear.And other children's stories about a wolf in sheep's clothing.https://t.co/TOxClT60EG
We live in a society.https://t.co/Pb3IYvA7IQ
Please submit all personal data to the advertisement regime.https://t.co/KXMCfm65Oh https://t.co/9kAkkRwf3i
You will be assimilated. Resistance is futile."The landlord has made it clear that tenants will not be permitted to opt out, [...] So there is no concept of consent in this—really, it’s the tenants having no agency over this unique information."https://t.co/4Gofgnanuc
Remember when we used to have a smart thermometer that could tell us how bad the fevers were for people getting sick (with the novel coronavirus going around); You could make better risk predictions about social spread? Boy those were the days, huh.https://t.co/stYx6c0SRT
Now I get a big meter that says "BAD" or "MORE BAD" that changes if I'm looking at Vermont as a whole, or just my region. Which one of these would you choose? https://t.co/j5fnU0TkWJ
Trick question, I'm gonna go by the bordering state (NY) that we'll see cars traveling from, which I can't see easily from the map.Wish I could actually see if people are having fevers or not. https://t.co/FRR3Hg6Fhe
This was such a better app in the before time! Ignore how the data changes between screenshots even tho we're looking at the same day. I'm sure that's fine and normal and not a sign of shoddy data management. https://t.co/XmVU98C4QL
I like how the expected and observed illness numbers change post-hoc, like they were backfilling data from other sources instead of just going by their own thermal readings.Does that bother anyone else? They never mentioned why. These aren't even available anymore.
The question is, do you think they're no longer using their internal data and so what we're seeing is a glitch in data gathering?Why is there a gap in Windsor county? https://t.co/uNDl42YBa2Suspiciously like they just didn't do reporting around thanksgiving. https://t.co/cHO4Gkk63N
So if kinsa is just reporting state numbers and ignoring their own thermometers on https://t.co/VJ8zwXYaik, this would explain the weird 'predictions'.But I can't tell! Maybe the https://t.co/3WTjaebcan having missing days and then a spike in data is just a coincidence? https://t.co/OmpJxsftSf
I like how nearly every Sunday we see a spike in cases in New York County. Is this an artifact of not reporting the samples over the weekend? Or does coronavirus take a vacation on the weekends? https://t.co/SqOYgcn8Zk
Were does https://t.co/3WTjaebcan get their data? https://t.co/6BnqkOz5TH
Two important facts about Vermont:Demographically speaking we are one of the older states.But we are also one of the most vaccinated.I really wanna know why these numbers are so high. https://t.co/hRTXMleqpF
Is it really just a spike from thanksgiving? https://t.co/uEdFscufrw
We had a "glitch" back in September where a downstream IT vendor failed to update data and it made the numbers spike.Given the reporting gap, a sudden trend from thanksgiving that may not be higher caseload isn't unreasonable.https://t.co/Pes8FjvMsf
So why is https://t.co/3WTjaebcan showing major spikes in vermont on the 29th and the 2nd?Is there lag time on the PCR data?(screenshot from directly viewing arcgis)https://t.co/hwHgoRacXQ https://t.co/PTQI01sPD7
It seems that NYtimes disagrees with the view from https://t.co/3WTjaebcan, only showing one spike on the 29th and none on the 2nd.https://t.co/LrIYLAma1B https://t.co/epDjll8lDL
According to the nytimes, the hospitalization rate hasn't really gone up in relation to the spike. It's high, but its not "suddenly we quintupled our caseload" high. https://t.co/p330RF1pLw
(threading broke, continue here: https://t.co/tGJBUZMOcl )
Sure, there are 600 case on december 2nd, but ther were also high number of cases on nov 4, 11, and 18th...Did we actually see a spike on thanksgiving ?News says we did. Are they wrong?https://t.co/QIp56KwwHR
Please don't lie to me when my family's health is involved.https://t.co/3sT5B99uTs
I sure wish there was some way to get access to reliable smart thermometer data.Would be really unfortunate if all we're seeing here is artifacts from false positive rates.https://t.co/bQ0ZMLL5Z5
*rolls up sleeves*https://t.co/0fLVXs8624
News is telling me there is a spike, data is telling me there is not anything other than a glitch from lack of reporting.Do I need to protect my mom again? Or is this just noise from increased testing leading to false positives?No idea. And I can't confirm anything.
New York on dataWeather https://t.co/uv5BUvvdfD reports 20,762 cases Nytimes https://t.co/J8wtv1mMj9 reports 11,331 casesstate's .gov website https://t.co/HkScE8NAuf reports 11,242 casesHow can each of these be different numbers?!? https://t.co/AxbkRYlwrM
I found the place where Vermont repots cases, it is also in disagreement with the NYT and the Weather https://t.co/J7NzUU0B8x
Found out that weather underground seems to show the same cumulative data as the state website's. Cool!https://t.co/C2BGzOekwj https://t.co/qtQT1YSARk
Compared to an earlier screenshot from Dec 3rd in this thread, the number of cases in Vermont has jumped to 2642, but the number of people in the ICU actually gone down by 4, and hospitalizations is only up +11https://t.co/Rh5w4CpqS5 https://t.co/oOa5TRB5yC
Guess we just have to wait and see."However, if there are still so many cases, the big question is "How many hospitalizations will that result in?" And that's something that we're keeping a close eye on."https://t.co/ZLClhUl52r
"Good news because that means fewer people need oxygen and ICU care. This also fits with some early lab research that shows omicron does not infect the lungs as well as other variants." https://t.co/8KQOHBnRKZ 🎉
"It is happening already; in some states, governors have lifted restrictions, allowing hair salons, nail salons and gyms to reopen, in defiance of warnings by public health officials that such steps are premature."https://t.co/kuOqSTwZCF
"The state had already been loosening many of its restrictions, allowing vaccinated persons to shop unmasked and dropping mask restrictions for all persons outdoors." https://t.co/nXE4d6qS3k
"Some jurisdictions reverted to mask mandates even before the CDC updated its guidance, citing the spread of the highly transmissible delta variant and increases in coronavirus cases, hospitalizations and deaths." https://t.co/uEwlkXYydJ
"We’re trying to get more people educated, we’re trying to bring them to us to get through this. There is a social science to this and sometimes forcing people, or potentially forcing someone, into something they don’t want to do hardens them" https://t.co/1AJ7OgTU0S
There are currently a 1/3rd the number of people in the hospital with covid, and only 5 in the ICU. Directly from our state's health data. https://t.co/hwHgoRacXQLowest it's been since I started looking at it regularly.https://t.co/rzNjs7Ijwm https://t.co/6K8lOVclvW
NYtimes is still reporting new cases on a daily basis that appears to be adding up the previous missing days and then using rolling averages. Those big spikes coincide with the missing cases the days prior.Do we not timestamp when samples are taken? https://t.co/LrIYLAma1B https://t.co/nxujE9lMFW
NYtimes hotspot data also is different than Vermont's reported case counts. It seems they are taking into account trends in case load... But I question that model given they have so many data anomalies. https://t.co/VRfdKtDESl
Both the State and the NYtimes data disagree with the CDC's recently released Community Level risk.https://t.co/T0Bl5QgGnFI do know that there is a big sign outside the local grocery that says the local municipality has said we are supposed to wear masks. https://t.co/JkyjS35MpX
If only there was a reliable source of community transmission that can be tracked using municipal measures. A population level sampling method that predicts Covid case by a couple of weeks (that I've known about since at least march 2020)...https://t.co/KrFY3qCCzm
Gosh, remember when I blocked the CDC's account on her for telling me not to wear a mask. Those were the days.https://t.co/tvNqawENix
"We must acknowledge in addition to changing knowledge about the virus that the “weather conditions” have changed – omicron is more contagious than delta, which itself was more contagious than earlier strains, thus requiring more protection"https://t.co/KmrO3N9OYT https://t.co/Y4l9M18NDC
> “I think it’s hard to process what’s actually happening right now,” said Janet Woodcock, acting commissioner of the Food and Drug Administration, during a Senate hearing Tuesday, “which is most people are going to get covid.”https://t.co/wYUpzabY43
I like that the CDC data on covid severity actually shows me what the larger pattern looks like and isn't restricting it to state-only now.https://t.co/9lQWqEygzo https://t.co/5qErW93RR2
seems to reflect local numbers a bit more now too.https://t.co/hwHgoRacXQ https://t.co/OwUBX6E1aj
Seeing levels spike in waste water and then showing up on the maps shortly after is comforting because it suggests I am mostly right about my ideas on modeling disease spread.https://t.co/2hFmWOo59KI like being able to poll the data directly.https://t.co/wWRXkehRal
"city was also primed for wastewater testing because the sewage system is split into three sections, each covering a specific geographic area. If the virus is detected at one of the plants, it can be located, at least initially, in that region of the city" https://t.co/BX7XPjelnr
It's important to take these per-county level readings with a grain of salt.Population density should be taken into account. Here's a 7 year old image compiled from "Census Block" https://t.co/YV36iwzgga
There's a bunch of these on /r/MapPorn/While they are out of date by years, they are good for eyeballing risk.This one also had an image of lights from space in the comments https://t.co/Rt8IaN8YNh https://t.co/hUL1aqHW6M
Hey, if we're going to spy on everyone using cellphone data, at least we can use it to map and monitor flows of people across the nation to guess at stochastic models of disease spread.https://t.co/eOLuV2yz9Z
Ever wonder if techniques behind stingray can be scaled up to the entire world?https://t.co/4W9DXvoOEb
I couldn't find direct data about movement within vermont, but with a bit of inference and imagination you can mentally pretend what it looks like.I extrapolated from this visualization of France's flows https://t.co/wbeenWkmdz
"are you aware at all of the current state of surveillance and what if anything has changed since your revelations?" https://t.co/tAI65GHw8F
News on a study about using cellphones to track movements during covid crisis.https://t.co/g4P9TTS4ij
Well that sucks.https://t.co/LgRCQdGNs8
Why has it taken so long to get analysis of this data visible to the public?All we get is a dashboard with "low, med, high" warnings that have no meaningful breakdown in how they are calculated. And yet the CDC has bought data from multiple brokers?https://t.co/UjIT7mmwZa
This is annoying. They're moving to weekly? Why isn't this some automated computer system that can constantly release data?https://t.co/YkCXRMQ8b3What is this "reporting burden"? Can we make it not a burden?( h/thttps://t.co/U9Sthz0Y6v ) https://t.co/VP533jUsmO
Please stop sucking.https://t.co/7utIJMQAcJ
Rutland county, why do you always suck so much?https://t.co/dA9IyNEkrK https://t.co/nqXGgpOVqd
This data seems bad. No deaths for a year in Rutland County? They test count is also really low so the % positive score is misleading. https://t.co/kfgauVUrrO
Is the Data bad? Here's the official chart from Vermont's report from January 25, 2023. Definitely shows death counts for that time period. But maybe it was from other counties.https://t.co/ELrVZtDg0Z https://t.co/7DagtbZk30
Their help content was updated today...https://t.co/O3ja2uez8b
Their videos remark that it is meant to help people plan around covid concerns. Kinda shitty if the data isn't accurate.https://t.co/rSRSP9T39n
Published today."The U.S. Still Doesn't Have Good COVID-19 Data. Here's Why That's a Problem"https://t.co/FfwRpGXsaDGarbage, its all garbage! Data is crap. https://t.co/lllnN5bwm6
Yay, we're all without knowledge again. Just like old times. https://t.co/H0uM1ctUUr
Our aging healthcare data tracking system has died suddenly it seems.https://t.co/7YplS95faR
Haha, and the NYT's data tracker says vermont has had some strange spike in deaths?Where did they get this data?https://t.co/LrIYLAmHR9 https://t.co/SrbgAEbHs8
0 deaths on Jan 7th, to 13 deaths on the 13th? This is a data gllitch. https://t.co/pGo44Yt7eK
These aren't new deaths! These charts are deceptive. https://t.co/8OSFHaUirn https://t.co/ZSjQPNSx2D
Dumb."The Department found that, due in part to the reduction in staff capacity as it scaled back from peak emergency operation, some death reports were not processed correctly, resulting in the reporting discrepancy." https://t.co/mcsophL5H6
LOL, now the CDC's website for tracking risk just says "undefined" everywhere. https://t.co/ClRF7Yg4ij
The CDC wastewater reporting map is pretty good. Almost all of New York is showing a rise since the beginning of September.You can also spot the ones that are fucking up reporting.../me *gives side-eye to Chittenden county*https://t.co/pfXI9qkmB8 https://t.co/rAUcj41uK2
The CDC's waste water map is now just state level and says "very high". This is absolutely useless to me. It doesn't tell me where the readings are from and with Vermont's dashboard shutdown I can't even see what counties might have more of it.https://t.co/da9sN7DAnj https://t.co/QXZaGwQ21K
Remember when we had good COVID surveillance data? Now we just get some weird chart. And if I click thru to the map it just links me to healthvermont[dot]gov with no additional data.Fucking useless.https://t.co/KSRUHOYgNV https://t.co/dfEugXtYDm
I don't care about how many people died, I care about infection rates. You fucking fucks.https://t.co/HCIp935xlO
Ok I finally found a link that shows actual wastewater measurements. Ironically it was the second one cuz the first one doesn't' tell me shit and 404s.Good thing I had the foresight to save the direct linkhttps://t.co/pfXI9qkUqGhttps://t.co/k4bwTEUkK0 https://t.co/QUx10cJpaM
New York Times' covid tracking charts haven't been updated since march and tell you to go look at the CDC, linking to the useless weird chart.https://t.co/32JZ9pD72S https://t.co/f3WOUVphNe
We need better disease surveillance.https://t.co/KOgBEZUUxu
I was promised Total Information Awareness and all I got was an aging heathcare informatics system being slowly farmed by greedy platforms propped up on VC driven growth fetishism.https://t.co/zr215dxJ7jhttps://t.co/LRycI9xQjR
The dual-use of monitoring for racism using large scale data tracking regimes (using the kind of data that the CDC hoovered up during early covid)... is not a good one.https://t.co/bo3iPB1LMn
"What could go wrong?"https://t.co/4FeoEhihTD
I'm not."the military could target rank-and-file militants as they were at home surrounded by relatives and neighbors, instead of only when they were alone outside." https://t.co/53yWUGbKgbhttps://t.co/Mk0ZW23IDq
While you are distracted by the security and privacy apps on your phone. I'm looking at how to avoid WiFi.Don't get distractedhttps://t.co/ahSyU52FDthttps://t.co/17M7bTlRIo
Hello! Did you know your phone can be tracked even if you install a VPN and Brave and other anti-surveillance capitalism tech?For 'heath' reasons.https://t.co/2tD49SJtVm
The thing I'm worried about this is this stuff being done entirely via the cellphone connection towers itself, but Those tracking services are also hoovering up location data. https://t.co/7572hjNQRF
There's tons of these out there. Entire networks of B2B APIs and backfilling tech to de-anonymize location data patterns, eventually offering this to government entities. https://t.co/7wf7eies0Z
You think that's a legitimate cell tower you are connecting to? How would you know?https://t.co/VtCaioDJvL
Wardriving for spooks.https://t.co/Ejy4Z66YS4
I wonder how long until someone is thrown in jail because an AI scanning their tweets for something construed as "intent". We're already killing people based on AI systems that scan social media it seems, so it's only a matter of time. https://t.co/mQS9c61Gb9
