2,593 karma · joined September 29, 2015
"Some users may be surprised if a biometric authentication suddenly appears on a website or an app and think this is sending sensitive information to the server. With passkeys, the user's biometric information is never revealed to the website or the app. Biometric material never leaves the user's personal device."
Having a notification asking me to enable Bluetooth immediately concerned me even though it was clearly the genuine prompt you get get when Google asks you to sign-in by accepting a prompt on a phone.
I don't see myself using this authentication over pass+2factor even though it should be safer because of 2 main reasons. I avoid turning or keeping Bluetooth on and I don't see a clear way to manually control and export the passkey private key. I understand that it's synced via the OS and Google's password manager, but I don't trust it unless I can back it up and restore it to a device myself.
Is there an open, Google independent implementation like Aegis is for 2 factor?
If I can't control and manage my private keys, it's not my private keys. The more I read about this, the less this feels like an improvement to pass+2 factor. Phishing attempts are easier because Banks don't allow non-sms tied 2 factor authentication.
Thinking of this like SSH key based authentication for website made the value of this clear for me. But I don't want my passkey dependent on my Google login.
Doing one will help you prepare to get the most out of the other. Living somewhere and doing immersive language learning would help you understand and enjoy media and food specific to that region. And watching shows and making dishes would help you connect with people from there and feel less like a tourist if you were to visit. It's depends on the individual whether the big things or the small things will push them to grow the most.
It was able to shift by a constant number successfully and even tried shifting both forward (+2) and backward (-2) looking for valid words without additional prompting. But it did not loop through every possibility and stopped having found a word that wasn't real. The interesting thing was that asking the model if the word it found was real with a follow-up question, it correctly identified that it gave an incorrect answer.
Part of why it failed to find a word is that it did an incorrect step going from EXXEG... to TAAAT... as a poor attempt of applying the frequency analysis. It understood that E shouldn't substitute with E and moved on to E->T, but the actual substitution failed.
The limitations of context memory and error checking are interesting and not something I expected from this model. The unprompted test of both positive and negative shift values shows some sort of system 2 thinking, but it's doesn't seem consistent.
I don't believe it's possible to integrate with modern society successfully without having to go through some kind of culture shock. It's not limited to culture defined by regions, older generations feel culture shock constantly when trying to stay integrated or understand the culture of new generations. But things like YouTube have allowed me to experience culture in a way that wasn't possible 20+ years ago.
I have been loving the relatively recent trend of older people, parents and grandparents making cooking videos. I recently found De mi Rancho a Tu Cocina, a channel where Doña Ángela, a 72 year old grandma teaches Mexican recipes. The channel has 4.2 million subscribers and the first video was uploaded in August 2019 and hit 2M subscribers by December 2019. Very simple production value, except the wood fire cooking which is probably out of necessity as she lives on a ranch. A similar channel with more production value, but still very homey is Cowboy Kent Rollins, 2.32M subs, who is verified on YouTube and has built a brand and a merch shop related to outdoor cooking. I grew up in the suburbs in the tristate area, but love making myself traditional style Mexican and TexMex food, homemade refried beans is a favorite.
https://www.youtube.com/@DemiRanchoaTuCocina
Best Authentic Refried Bean Recipe - Cowboy Kent Rollins https://www.youtube.com/watch?v=iC5mR5s70bE
I grew up 30 minutes from a respected state college, about 15 miles, and about an hour away from both Philly and NYC, almost exactly 50 miles to each. The ongoing joke in my high school is at least 50% of each graduating classes ended up at the same state college. I'm not sure the real statistic, but I don't think there's anything wrong with that.
10 miles sounds far enough away from parents to be able to feel independent but also not far away enough to feel any kind of culture shock. Not everyone will have this privilege, but I just want to share another perspective.
There's a very clear distinction between local and cloud, every single photo taken using any camera app is local only. Galleries, like Google Photos, and backup services like Dropbox have an explicit setting to enable cloud backup. Google Photo backup is very distinctly different from Google Drive and phone backups in general.
I use FolderSync to sync it with a self hosted NextCloud instance. If I am travelling, I prefer using Syncthing, a wifi file server running off the phone, or a physical wire to transfer photos. I know exactly what pictures are sent where and backed up with what method. I am not sure why you believe Android treats photo and file management like a black box, I don't believe Apple does either. Apple iOS devices are much more difficult to manually backup as they only allow background photo sync to iCloud, if the screen is turned off syncing to a third party will most likely fail. But there is a straightforward toggle to disable Photos backup to iCloud.
Android file management is superior to any other mobile OS. Android is file privacy first by design, as long as you are comfortable managing backup and syncing with self-hosted and self-managed solutions. I have gone the additional step of rooting my phone as well to get around any Android limitations on what folders certain apps have access to.
No one is arguing against the last point you made. No one was conflating the two.
What stood out to me what one of the first couple comments in the support thread was the contact information for the Museum Director of the Swedish Tank Museum.
My unconfirmed theory is that Google's OCR pdf service flagged specific text and pages in the pdf of tank plans and repair manuals as they are considered classified and shouldn't be in the public unredacted. It's historically significant and absolutely worth preserving, but I can see why it might get flagged.
This was also interesting in that it got the entire account taken down. Usually Google flags a file and disables sharing completely. Google disabled sharing backups of a recent Kanye interview that was filled with hate speech called "Kanye champs removed video.mp4". I have not watched it but it was my first time seeing something removed because it "may violate Google Drive's Hate Speech Policy. Some features related to this file were restricted."
I am not sure if it was the post to HN or various videos other made that helped, but it made me realize that the one of the best ways to get in touch with a human at Google is posting here on HN and that Google will continue to become more aggressive with scanning files.
Section 230 was modified by FOSTA-SESTA when related to hosting sexual solicitation, and EARN IT is back under review by congress which would effectively make E2EE illegal and make hosts liable for illegal content uploaded by its users.
Would scanning be okay if there is a government entity with a human you can appeal to that would override any flags made by automated systems? No corporation wants more government oversight, and the only way to avoid it is to do good faith self policing. EARN IT would receive a lot more support if the public sentiment changes to believe Google and Apple and other hosts intentionally choose to ignore problematic materials. Reddit is criticized often because it allowed problematic subreddits to grow. I am not sure where things will end up.
From one perspective, I understand that cloud storage should include certain rights to privacy. If I buy some compute time on AWS, I manage and control all flow of data. It would be a complete violation if AWS policed the kind of content I could store. I think self-managed vs company managed is what dictates my expectations. Google photos and drive, especially since most people don't pay for it, fall into the company managed category in my opinion.
In the last few years Google Drive became more strict about sharing copyright work and started flagging files, which limits the ability the "share to anyone with a link". For people who use it as a backup, the advice I started hearing was to encrypt the data before uploading. The opsec while considering nation-state level threats to privacy already recommended only using encrypted containers since the beginnings of cloud storage.
The Communications Decency Act is an interesting law that dictates the liability of online content and service providers. There is an ongoing battle to increase ISP and host liability as a way for the entertainment industry to combat piracy.
FOSTA-SESTA and the ruling on Backpage [1] made things more difficult for hosts like Google and changed the liability surrounding sexual solicitation. If a host like Google knows what's being hosted and does nothing, they get in trouble. They don't have the protections they did in the past.
Not having a human in the loop is terrible and a problem frequently when it comes to getting support from Google companies. That's a separate issue from Google or Dropbox being able to scan unencrypted files for CSAM. Google's policy of automating as much as it can has tons of downsides. But it's understandable when you look at the scale Google functions.
It's important to separate the policy of scanning from Google's terrible appeal process and the algorithm false positives.
I would feel differently if the story was that Android scanned an outgoing SMS or a photo saved locally. I am not sure where the balance point is to identify and report CSAM while also respecting user's rights to privacy.
I am glad the more blatantly privacy violating policy was canceled, but I can't help but be cynical. Especially as Apple has "followed local laws" when strong armed by certain governments, and a recent class action was filed for lying about if Apple tracks you ignoring explicit settings not to.
https://www.apple.com/child-safety/
I can't imagine people working on E2EE at Apple would be okay with client side scanning. The reasoning isn't important, it's an easy slippery slope once implemented. I imagine the encryption team has to constantly push for consumer privacy in a climate where privacy is challenged and compromised for ad companies and governments. I would be absolutely shocked if there wasn't a large amount of internal pushback when the old CSAM detection plan was first announced publicly.
"When receiving this type of content, the photo will be blurred and the child will be warned, presented with helpful resources, and reassured it is okay if they do not want to view this photo. Similar protections are available if a child attempts to send photos that contain nudity. In both cases, children are given the option to message someone they trust for help if they choose.
Messages analyzes image attachments and determines if a photo contains nudity, while maintaining the end-to-end encryption of the messages. The feature is designed so that no indication of the detection of nudity ever leaves the device. Apple does not get access to the messages, and no notifications are sent to the parent or anyone else."
Apple, the client side scan pushing and ad platform expanding company is now the same company that is releasing strengthened cloud data protection. Deduplication becomes impossible at any sort of scale and for safety Apple even turns off web access to iCloud when E2E cloud protection is turned on for the first time.
Apple has stated it will cache thumbnails using standard protections when sharing files, using "anyone with a link" will expose the unencrypted data to Apple servers. I wonder if CSAM scanning can take place for those files only.
Open alternatives, like GPT2, have already been released under an MIT license. Community language model trained on NSFW text already exist as well. In my opinion, a move like this would only push away most people in the long term. Given the choice between a non-censored good enough model, and a censored great model, I would choose the good enough model for 99% of my long term projects.
If I am really struggling, I play chess puzzles or 3+2 time control on Lichess. I haven't learned any meaningful opening theory and I don't care to right now, tactics or the Chess960 format with a computer is nice for that.
After gifting technology to some family members, I realized that what matters most at the end of the day is they enjoy using technology stress free. You could get away with non M1 MacBook Pros for presentations, but if using an M1 makes you feel good then you should use it.
If you're worried about e-waste, then pass down old technology. My sibling got a phone a lot younger than her peers because I love technology and when I switched to a new model, I passed it down. She didn't care about the specs, like 99% of the population she doesn't think about compute power.
Compute power isn't a limited resource in the world. If someone has a need for compute power, they will probably find a way to use old technology on their own. That's not something you should feel bad or worry about. I don't.
I know that Brother includes a starter cartridge that usually includes less than even a standard size replacement. They have high yield cartridges and I have only ever seen the bigger size in third party replacements.
The model I have is the 2270dw which was purchased years and years ago. The bigger replacements are rated at 2600 pages and standard are 1200. The drum unit can do up to 12000 and I don't think we've replaced that yet. I did reset it a couple times to get past the toner warning, it continued printing well into faded pages without issue. All it took was holding a button on the front while turning it on and then pressing the same button a specific number of times after initializing. The third party replacements are priced reasonably well in my opinion.
I am not familiar with the OpenAI workflow, but KoboldAI is a community open source implementation to interface with different AI language models locally or online. There are links to premade notebooks of open sourced models you can run different on Google Collab. [1] I think finetuning models offline is still too expensive, but there is some progress and is possible via collab.
KoboldAI has a chat mode setting where it properly generates and continues a conversation between two people. The models I ran were the smaller ones, but there's larger models available on huggingface like FB's OPT 30B. Even the smaller models I could fit into 8gb vram were coherent enough to be impressive.
I am surprised I haven't seen more of is a chat bot based on writes who have passed away. I imagine an Aristotle chat bot would be pretty interesting and avoid issues of copyright.
I also imagined that muscle recovery would be less efficient or take longer from a fasted state. I have hit rowing/ERG goals and made time improvements so I feel comfortable saying that fasting doesn't hinder my cardio.
I understand that energy goals are correlated with body size, we were discussing a general minimum kcal threshold for unsupervised dieting. I am unsure why you feel estimated BMR/RMR - 200 kcal is unscientific. I was told that without supervision, it would be hard to ensure proper nutrition on a very low calorie diet. Additionally, he expected that some people would fall into a binge-restrict cycle because they can't sustain their usual energy level.
I was additionally curious because there have been days I forget to eat, and thinking back I realized my total calories for those 24 hours totaled less than 600 kcal.
I think something like Tex-Mex would easily allow for clean eating. Avocado toast, with whole grain bread is fairly healthy and was an absolute favorite of mine after workouts and I would consider it "American" food. Tacos are a great way to have good portion control and also feel filled. Allowing non-traditional recipes opens many options. Avocado grapefruit salad on tortillas with fried tortilla strips is a recipe I came across recently that I really enjoyed. I think non-traditional is the essence of true "American" food.
For fast food, Popeye's Blackened tenders are not breaded, a 5 piece serving is listed at 280 kcal compared to the classic tenders being 740 kcal. Nuts or apple slices with peanut butter is a fairly filling snack and can easily fit within 200 kcal. The main thing you would have to look out for is sodium if cooking at home is difficult. I know most of my bodybuilder friends go to a diet of chicken breast and brown rice for dinner for months at a time when they want to cut. It's around 400-500 calories depending on how you prepared it.
Vegetarian and vegan recipes are good starting points for cleaner eating too. Borrowing from other cuisine makes it a lot easier as well. I love preparing vegetable side dishes for dinner in the way it's prepared for Korean dishes. Miso soup is fantastic for when I have food cravings late in the night. It's not "American" but I'd argue there's nothing more American than making microwave miso soup at 3 am in my star wars mug. Coffee is popular too, and is very low calorie if you can get used to no milk or sugar, or only sugar alcohol/substitutes.
For some context, I grew up and live in one of the most culturally diverse states and counties in the US. I know that not everywhere in the US has the same diversity of food options, but I would argue that Amazon has done a lot to make food accessible.
We both come from a South Asian background and discussed how the number would change for a home cooked vegetarian diet which is more nutrient dense but low in protein, and I was told that would lower it below 1200, but he didn't feel comfortable recommending that diet unsupervised. He emphasized that I should decrease calories by removing carbs, then fats, then proteins. And that I should prioritize eating proteins above fats and carbs, especially if I planned on increasing the amount of exercise I did.
I asked him if the 1200kcal was net or gross. For example if I did 2 hours of moderate exercise, should I eat back 400kcal+ to maintain a net 1200kcal. He didn't have an answer for me, other than saying, if I start to feel weak or lightheaded I should immediately stop exercising. I don't have pre-existing conditions related to blood glucose, but I imagine that would would change many of these guidelines. I was told to stay mindful of hydration.
The stories I hear of fasted work outs is the reason I asked him about low calorie diets. My non professional intuition makes me believe that even intense cardio should be okay, but moderate or heavy resistance training should be avoided on a low calorie diet.
The general advice, even from my primary care physician, has been that anything less than 1200kcal/day is dangerous and is crash dieting. Looking around, I found this other study studying total diet replacement providing 810kcal/day as the sole food for 12 weeks, showed significantly greater weight loss compared to standard counseling and modest energy restrictions. The difference in weight loss was measured at 12 months. (BMJ 2018;362:k3760 [1])
25 BMI, which the diabetes study used in addition to T2D, is not considered clinically obese in the US. 25 BMI is the line between clinically normal and overweight. I wonder if crash dieting is advised against on the assumption that most people will go back to unhealthy eating habits instead of changing their diet, and also become more likely to fall into binge-restrict cycles. It's easy to imagine that if someone is able to make meaningful changes, crash dieting is more effective than most other methods.
The other issue is that the AI weighs weak openings way very heavily for both elo and responses. I enjoy chess960 and am weak on my openings, my elo is estimated to be between 1000 and 1400 standard otb 10+5, and my lichess puzzle elo is 1600. Noctie gave me an elo of 500ish even when I won. The response moves weren't challenging and it felt like the game went directly to endgame after a series of aggressive exchanges. Look forward to seeing where this goes.