scrps

People having a visceral hate of advertising and a vast swath of the internet's users block them in every permutation because advertising is good for ones mental health?

Car ads selling sex appeal, power, status and freedom are provable features of the car? Same goes for influencers pushing products, submarine marketing, etc... It is all psychology and it is manipulation at scale. Words at scale designed to manipulate people into doing something is psychological.

This is an interesting read:

https://blog.ibwave.com/a-closer-look-at-attenuation-across-...

I am tempted to disolve a bit of the plaster and run a conductivity test on it for fun. It is like living in a leaky faraday cage.

Here is where it gets interesting... No mesh, it is lath and plaster. I've had to cut into it in a few rooms and if there is a mesh I have yet to see it. I am wondering what was used in the plaster at this point.

The one upside to this being based on open standards also means that there is no reason this tech can't be used against say, police, politicians, executives.

Though I'd like to actually look at how it is practically done first, for all I know this could require more than just access to an AP. This is not my area of expertise so maybe someone that is an expert can confirm my intuition that beamforming even on enterprise grade access points isn't locative (if that is the right word?) enough on it's own, I'd imagine this would require some type of kit to be deployed in order to accurately map a room?

Also plaster walls or at least my plaster walls seem to be great at blocking anything over 2.4GHz :\

I completely missed that somehow, thanks.

Wouldn't you need more than that per person for redundancy? I assume that would help protect against some forms of crop loss?

Turning a critical peice of internet infrastructure into a financial toy for wall st would be a bit akin to taking all the physical barriers to access off of the nuclear football, authenticating it, and leaving it in a room full of bored children.

Don't do that.

I am just using free tier right now but I think the privacy bit is a sweetener, the value with kagi so far for me is you can scope searches really well, add preferred sources, filtering. It's a bit like if you slapped a nice interface over all the google hacks that used to be possible. Lack of SEO junk is another sweetener.

Yeah, a lot of this is technically possible without kagi but if kagi doesn't suck I'll pay to not have to deal with the extra cognative load of wrangling google or ddg into spitting out sensible results.

Well that finally explains all the unsigned birthday cards that play a midi version of Every breath you take by The Police I've gotten.

Another week, another rugpull.

Thank you

I use kanboard as my personal board, it's great software, but also tooled more for orgs and this seems much more oriented toward my needs so I will absolutely give this a spin.

Hey no problem. And if you feel like you are really having a hard time finding your way talk to a professional.

As for the degree, you got one, sure it wasn't what you enrolled for but you got a job doing CS WITHOUT the degree, sounds like your attempt was a success not a failure despite a setback.

There are thankfully, quite a few, ive mostly used rocm/rocm-terminal and rocm/rocm-dev.

https://hub.docker.com/u/rocm

Not a coward.

I can't tell you the number of times I've felt like I know absolutely nothing, that is ok and I've met a lot of people who go through it even if they've been doing this for 20+ years. What you feel as "knowing nothing" is the universe telling you the scope of what you are trying to master, it is meant to humble you, don't let it crumble you.

As for doing things from scratch, that isn't as easy as it has been in the past, look for a project doing something that speaks to what made you get into computers and engineering and see what you can contribute, sometimes the money aspect of this industry is what kills it's soul.

Take breaks, give your brain time to rest and get perspective, do things that are far, far away from anything to do with engineering and don't forget absolutely none of this at the end of the day is ever worth harming yourself over.

We are running a very similar setup with the exception that we went with Alligator Valet... Conceptually very similar but we found it snappier in production.

I would hope both would set off alarm bells but I guess at least one of them inducing trepidation is cause enough these days for guarded optimism.

This is a bit batty and not sure it would work but I wonder if you could expose /dev/hidraw using sshfs then your work machine would see it as a local yubikey.

Spotify's rating and recommendation system leaves quite a bit to be desired. They need granularity, and they need to explain the algorithm to some extent.

If I like a track what exactly am I liking? The genre? The artist? Only specifically that song? Some maths that approximate a certain properties in the track? Some combination? How does liking song X influence Spotify's recommendation system on average? I have experienced the same as other posters WRT listening to a genre for a session only to have Spotify fixate on that genre until I've skewed the average in another direction. Small hack I've found if you make a playlist with a large and diverse sample size you can nudge their playlist enhancer to stay with the mood you are going for.

On granularity, I would want to be able to say:

Let x = track, genre, artist

I don't like X, never play it.

Exclude X just for this listening session.

Skew toward more X

I'd also imagine this would produce much richer data for future recommendation systems. If Spotify were to train on the data I've generated on their platform it would end introducing bias by the nature of me working around their systems and still not getting what I want, granted I may be an outlier.

Edit: Formatting & typo

I'll add to the list: Dolph Lundgren won a Fulbright to MIT and has a masters in chemical engineering.

Same techniques (largely derived from psychology, statistics, sociology, etc) used in social media marketing, intent tracking, etc has uses in national security. Same thing that might let meta know you might be interested in batman pajamas can be used to sort out if you are being radicalized to action by some group. Interesting example is fractalrabbit.

Group of people who both have the theoretical training and practical experience is probably pretty small so private/public crossover isn't unreasonable.

Thanks! The Alto restore was great, I ended up binge-watching the entire thing in one go.

I think for big orgs there could be a two-man rule system like they use in minuteman missile control, two designated members have to both agree to delete a repo or anything seriously irreversible.

Sausage is a good analogy. It is both (at least with chains of trust) the manufacturer and the buyer that benefits but at different layers of abstraction.

Think of sausage(ML model), made up of constituent parts(weights, datasets, etc) put through various processes(training, tuning), end of the day, all you the consumer cares about is the product won't kill you at a bare minimum(it isn't giving you dodgy outputs). In the US there is the USDA(TPM) which quite literally stations someone(this software, assuming I am grokking it right) from the ranch to the sausage factory(parts and processes) at every step of the way to watch(hash) for any hijinks(someone poisons the well), or just genuine human error(gets trained due to a bug on old weights) in the stages and stops to correct the error and find the cause and allows you traceability.

The consumer enjoys the benefit of the process because they simply have to trust the USDA, the USDA can verify by having someone trusted checking at each stage of the process.

Ironically that system exists in the US because meatpacking plants did all manner of dodgy things like add adulterants so the US congress forced them to be inspected.

I'd suggest perhaps donating if it is awesome.

If you can't donate there are lots of ways to help: add a feature, host a pkg repo mirror, clean up docs, fix bugs, add/manage a package, donate older hardware for compat testing, help out with community support, github issues. Telling people it is awesome also helps but maybe a more detailed why would go a bit further to promote it.

Would be more helpful though if corp OSS users who have their entire operations underpinned by financially undersupported OSS projects stepped up to help.

Thank you, I didn't even notice that. Too late for me to edit the post :(

If my reading of it is correct this is similar to something like a trusted bootchain where every step is cryptographically verified against the chain and the components.

In plain english the final model you load and all the components used to generate that model can be cryptographically verified back to whomever trained it and if any part of that chain can't be verified alarm bells go off, things fail, etc.

Someone please correct me if my understanding is off.

Edit: typo

Ken Sherriff (righto.com) who's chip teardowns are often posted on HN is also a frequent guest on curious marc's channel.

Just don't apply ML as a solution?

Even without the bias issues, if I were a hiring manager how do I verify that the weights "MagicHireAI SaaS Corp" are using aren't leaving swaths of talent on the table that I don't even know exist?