- > Wouldn't you want to look at protein per calorie or EAA per calorie?
PDCAAS is a reasonable standard to use, pro-rated against total calories.
https://en.m.wikipedia.org/wiki/Protein_Digestibility_Correc...
- I remember another post that was very well-received where an individual hacker wrote his own homebrew iMessage client for his own personal purposes. HN really liked that!
I think HN exists at an intersection of individual hackerism and business. If a project is clearly by-hackers-for-hackers it gets a lot more leeway for unsustainable concepts / implementations. But this is building a business on adversarial interoperability, and many people who LOVE the concept and technical achievements will still post mostly critical things about the business model because it’s fairly clearly a very very challenging business model.
- Maybe the “right” grayhat/blackhat way to handle it is to use high-quality, convincing sock puppet accounts to manufacture consensus against the “conspiracy theorists”. It’s not ethical but its the more effective alternative if you’re already at the point of locking threads where people continue to point out that you still haven't fixed the problem.
- I still dont think its normal to resort to violence just because someone will only make $400,000 this year instead of $4 million as a result of whistleblowing. Or even no change to their income but their company will make less profit as a result of whistleblowing. Or they’ll “be embarassed” as in the case of eBay.
That’s not a “threat” - they’re in no danger.
- I don’t think it’s normal human nature to assault / murder / psychologically torture / ruin the life of / etc someone who points out what your group is doing wrong. It may happen from time to time, enough that it should be a potential expected response. But just like psychopathy and schizophrenia are abnormal, so is murdering or ruining the life of a whistleblower.
1-2% of the population may be a sociopath / psychopath — but its still considered “abnormal psychology”.
If someone had proof that a device I made was hurting people, I wouldn’t try to destroy their life or kill them.
A lot of this whistleblowing doesnt even have jailtime as a consequence to those who failed their duty of care - often it just means they’ll make a few million less dollars but still be plenty comfortable.
We shouldn’t feel its “normal” to murder / torture / assault or ruin the lives of these whistleblowers any more than we think sociopaths are “normal”.
https://arstechnica.com/tech-policy/2020/06/ebay-execs-sent-... <- this is not just normal “human nature”. It’s the result of abnormal psychology.
- I love all his videos. But this one is amazingly mind-expanding. Thanks to his previous videos and a fluids engineering background, I had the right intuition for what would occur...but holy shit his experimental method was so far beyond what I expected. I kept wondering how he was going to show us the level of detail he was teasing in the title and introduction, and seeing the resulting visualization was just so incredible and inspirational.
Really he may be the best technical YouTuber for true nerds who don't want things dumbed-down but still want it to be reasonably accessible.
- Digital simplifies the design a great deal. Analog you have to worry about so, so many things. With digital, as long as the levels aren't too out of whack, you can "just" focus on the theoretical logic and ignore most real-world effects.
I say "just" not only because the theoretical logic is still fiendishly complex, but also because there are still real-world effects and flaws in your components which both rear their ugly head often (all abstractions are leaky). You can never completely ignore the analog world, but digital design is almost always much, much simpler than analog for anything beyond rudimentary levels of complexity.
Obviously if you're designing the silicon for digital components, you care very much about the analog reality of signals. But those silicon wizards are the ones who are building the digital abstraction of the analog world so that the rest of us can "safely" ignore it, mostly.
- > The design of MLX is inspired by frameworks like PyTorch, Jax, and ArrayFire. A noteable difference from these frameworks and MLX is the unified memory model. Arrays in MLX live in shared memory. Operations on MLX arrays can be performed on any of the supported device types without performing data copies. Currently supported device types are the CPU and GPU.
Weird and unfortunate that a framework made by Apple for Apple Silicon doesn't support targeting the Apple Neural Engine.
- Training data for a medical diagnosis model would likely include enough info to de-anonymize the info for some participants (age, sex, zip code, descriptions). I'm not sure what the answer should be but I'm uncomfortable with the medical training data being provided freely to the world.
- > are you using anxiety with a bit of hyperbole here, or is it really making you anxious?
Depends on the setting. Sometimes they're for things that have greater than usual effects, like getting locked into a subscription payment or agreeing to additional charges for Spirit Airline flights. Other times they're for things that I have difficulty taking the initiative to address, like notification/unsubscribe settings, for which my ADHD only allows me occasional "windows" of time where I'm able to initiate or follow-through on changing the settings.
Sometimes it doesn't really matter and I just scoff at the bad UI and it's not a problem.
- > Michigan Tech University, which is the mining industry's MIT
ehhh...Colorado School of Mines is a much more significant powerhouse in the mining industry. I say this as an MTU alum.
> I am surprised the state of Michigan hasn't found a way to help these guys.
Not really surprised, Upper Peninsula doesn't afford many political connections. Families who go to that school aren't well-connected and typically earn quite a bit less than families at University of Michigan. Distance matters too -- it's a very, very long drive so there's not a lot of cross-pollination between the political centers and population centers and the Keweenaw Peninsula.
It's isolated enough from the rest of Michigan that the McDonald's by Michigan Tech is an official supporter of the Green Bay Packers, not Detroit Lions. Wisconsin and Minnesota students pay in-state tuition there. It's really not "Michigan", culturally. It's also very sparsely populated -- the nearest Wendy's is a two-hour drive away.
- Google "Tesla Diversion Team". They've institutionalized management of their fraud.
- If they wanted different employees they could pay to attract “real” talent. But jobs that pay $130k in private industry pay $80k at Universities. Honestly the guys I work with are very good at creating stable, long-living solutions with minimal maintenance requirements. They're not ambitious but they’re careful to create solutions which don’t have technical debt.
Management OTOH often shove poor solutions down the pipe after meeting with vendors.
- Huh? 3 weeks vacation, $3,000 deductible and $8,000 OOP max for a couple. Not particularly notable, and higher than most OOP maxes I've had at private companies.
I mean, sure I can go to some concerts for free at the music school and use the makerspaces but that really doesnt make up for the $50k+/year I’m leaving on the table.
You are incredibly rude for someone who is also incredibly wrong. It is strange that whenever we are one of those, we all seem far more likely to be the other as well.
Only the last two pages before the appendix is "the rule-making document", and the 4 pages of appendix A - just six pages in total. The rest is a dialogue on why the rules are needed and provide context to understand the intent of the rules. The rule starts at "X. ORDERING CLAUSES" on page 394 and is less than 2 pages long in total. It will also be necessary to fill in references made to "Appendix A" which is an additional 4 pages (397-401).
It's not surprising to me that both you and the other poster couldn't figure this out -- it's very easy to miss a section so small when it's titled similarly to sections like "IV. ORDER: FORBEARANCE FOR BROADBAND INTERNET ACCESS SERVICES" which are mostly discussion. That contains language like:
> Petitioners ask that the Commission reverse, vacate, or withdraw the RIF Remand Order, and request that the Commission initiate a new rulemaking to reclassify BIAS as a Title II service and reinstate the open Internet conduct rules. Collectively, petitioners make several procedural arguments for why the Commission should reconsider the RIF Remand Order. Common Cause et al. and Public Knowledge each assert that procedural deficiencies in the process the Commission used to adopt the RIF Remand Order are cause for reconsideration. Common Cause et al. argue that because the Commission failed to open the record to receive comment on the impact of the COVID-19 pandemic, it failed to adequately consider harms of reclassifying BIAS as a Title I service on public safety, pole attachments, and the Lifeline program.
Which is clearly not an order - it is a discussion with a goal towards justifying parts of the order.
There are also only 434 pages. Not anywhere close to "693". It would be very rude of me to point out that you might be "unable to read past the table of contents". To the contrary, I understand that it's easy to misinterpret the indexing of the table of contents as pages rather than sections, and I have empathy for someone making that mistake, even if it does demonstrate that someone probably hasn't tried to use the table of contents to actually read the document.