throwaway914

At my org -

I've been using it to counter disappointing performance reviews in a professional way that aligns with policy to "further develop the employee". I've use it to create first drafts of proposals for changing business processes (saved me hours of time just outlining an argument). I've uploaded most of our PDFs and had it do some "fuzzing" in a sense to find logic errors, or slightly diverging interpretations of policy terms. This has been used to feedback to our policy and planning folks to cover a hole, or clarify some random document. I've definitely used AI to respond to a bad coworker or supervisor when I'm depressed I have to suffer under/around them. I still proofread the mistakes, but I make more mistakes when I'm trying to sound professional and just depressed about the work. Took this to a dark place, but there's more success than papering over BS. I've helped get 4 new benefits for our employee type with my proposals and the supporting "evidence" the LLM can drag up and justify in an opening brief. The type of employee I am is expected to receive 22% higher pay this year.

I used to think I don't have the bandwidth or headspace for understanding our collection of policy documents. There are rare experts. LLMs have lowered the barrier to entry for m e to contribute. I wish I could be specific, but I helped uncover that certain benefits were not available to the type of employee I am, and how previous pay comparisons were done assuming something about total compensation. Well, my employee doesn't have benefits so the numbers they were comparing against were not representative of us. Got a proposal in there, an got a few executives to say "ohh".

It's been fun.

Hmm. I will say from past experiences I've only found the 'valley girl' accent to be in certain parts of California, and not representative of Californians. Growing up I remember when my sister and her friends adopted this style of speaking for a time. Usually at school, this was (from my perception) to project an air of being disinterested in bored in conversation toward authority figures. I also found it when someone wasn't capable of demonstrating a deep knowledge of something (where the bias comes into effect). I know it's incorrect, but I have many friends that think Californians don't have an innate accent and I've often thought this about how I sound when going elsewhere.

I think your comment makes sense, and I appreciate the way you detailed this.

I work in a very large organization with 50,000 people. It's painful and discouraging listening to the person at the top, or their close deputies, because the documents they publish for mission goals are so abstract and boundless. Nothing I want ever seems directly addressed in these documents that get passed to all of us.

However, I've learned that the middle managers below C suite are looking at these objectives and justifying their own plans in alignment with this. Whatever the people who actually "do" want to do in 2025, must be in support of this abstract crap that frustrates me.

I've had more success in the last 2 years aligning my wants and interests with those of my middle managing-leadership, as they orient toward the executives.

Executives are often trying to enable with vapid bullshit, and really it just shows they shouldn't talk to the people who "do". I think a better way of approaching the workforce and inspiring loyalty to the org, is to trust a deputy to read you into their particular program area so the high-level executive can come in and speak "to you" at an unexpected/encouraging moment. I've seen that done well, but not all hands messaging - ever.

(All blanket statements have exceptional circumstances where they should be disregarded.)

If you sound disinterested and don't articulate your speech for the technical audience you have -> you deserve the perception you get.

Biases exist for a reason. If I'm not hearing the detail I'm looking for or the sincerity I can trust, I know I'm not your audience.

Why would I blame a bank when it's easy to impersonate me?

There are root-cause issues here when identifying a US citizen, and US law enforcement.

I watched this, and the commentary from the 2nd tour guide seems non-serious. I did think the "valley girl" accent is what threw me off, but I stopped watching at "super cool sled". I just don't understand who their audience is with this type of commentary.

Why are we blaming the companies because the police are easy to impersonate?

This is very cool. I remember I got sucked into things like Ramda going down this functional programming rabbit hole :-)

https://ramdajs.com/

I think I'm misunderstanding what this can be used for, but something I'd love to see become "perfect" is preserving a GUI application state across reboots/logins/etc. Especially with flatpak GUI apps.

I'm equating this to condition-less coding, where you're trying to eliminate branching logic with bitwise operators or other math.

I appreciate you making this distinction; it's one I haven't understood clearly. My outside impression is that right-to-work is a campaign to smear unions, though I understand in some areas/professions they can only hire union workers.

I'm not saying it's practical for everyone, but while the default is LinkedIn I have found great opportunities without it.

Trading 1 trash platform for another. LinkedIn could be such a great place if it didn't abuse your information and subscribe to such dark patterns. I still remember having my entire address book harvested and spammed (way back when). I will never forget that. I will never trust LinkedIn. Doesn't matter what they do.

I maintain a daily email to myself in a shared mailbox. I have 3 sections:

SIGNIFICANT

  \* topic1

    \* this happened

  \* topic2

DONE

  \* This project

    \* Made this change

    \* Made another change

  \* Some training

    \* Completed this section

TODO ...

I'm in a setting where I'm incredibly temporary. I could be tasked elsewhere tomorrow. Every day I reply to my previous email and work on the draft throughout the day as my notebook. At the end of the day I send it, received in the mailbox I'm attending. I title the email "Captain's Log" and my supervisor and peers can read it, as well as the draft, whenever. This keeps them clued in on where my head is at, what I'm working on, etc. Great for performance reviews mostly. Not as convenient as something like my Remarkable tablet.

Neat date format :)

I use these for work within SharePoint lists:

"2023 Q3 August"

"2023 Q3 08/02"

"2023 Q3 Monthly"

"2023 Q2"

"2023"

"Weekly"

"Monthly"

"Annually"

I had an issue scheduling things for 1-off occurrences, and reoccurring. A lot of what we do is scheduled around quarters or months. This was my "best of both" because I can store the date as text, sort alphabetically, and the dated-1-off things sort correctly. The recurring stuff like weekly/monthly/annually naturally group at the bottom or top depending on the A-Z direction. It generally keeps the format of year-quarter-month/day.

Thanks for the heads-up about the NeoSmart pens! I have a remarkable 2.

That is a good story :) It makes me wonder what future Wayland protocol may enable clients/apps to advertise themselves as screen-reader capable.

Not to draw seriously messed up parallels:

I frequently get in arguments with my extended family because I believe pedophiles and criminals have rights. I talk about the documented history of innocent people being convicted of a crimes, but there are many guilty offenders. I want prison conditions to be better, and I use it as an example of "if you care about human rights, you care about them for everyone. Even convicted rapists and murderers." Sometimes I go off on another tangent because we should care about rights for immigrants as well. :) Then I lead into a 2nd bit where I say, "There are many pedophiles out there. People who have the tendencies, and people who have acted. We know that women who are pedophiles go under or un-reported." To treat some of this, we need mental health facilities available even for people we find disgusting and monstrous. We need a safe environment where they can go in anonymously and seek help. It may prevent someone who is just thinking about children from touching a child. Most people would rather just execute a SUSPECTED pedophile. That's how you get someone to hate themselves and the world and children and act on their impulses. Make the stigma and the reality for them worse. We love doing it in prisons. We love making prisoners suffer as a society.

So to draw the awful parallel: We need mental health facilities available for pilots with no ability to report-out on how many are effected. We need to ensure anonymity, and provide resources with no idea of how well they may be helping someone on the edge. Maybe loose/aggregate data points can be reported, but my strongest emphasis would be on providing anonymous support that is always and frequently available.

If you care about pilots and the passengers they ensure safety for, you need to extend the resources with no indication of how effective that is. You want them to receive care/treatment, for what very well could be a moments' ailment. If there is a long-term issue, it may help them with coping strategies. The alternative is letting it go untreated, and putting others at risk.

Something I would love to find is a practical/succinct guide on Linux kernel performance tuning. I'd love it to give examples of sysctl's you'd adjust for specific use cases like:

- a real-time kernel

- a single user mode kernel

- an embedded device kernel (like an esp 32 or rpi)

- general purpose desktop use (surely there are gems in here)

- to use the linux kernel as a hypervisor hosting others

- tuning the linux kernel for within a vm (guest to another hypervisor)

- tuning for gaming performance

I myself do not know enough about sysctls, and I'm sure it's a goldmine.

I often feel this is what's happening: Portland and a few other cities around the nation prioritize helping folks in a big way, and we wind up serving a national need not a city need.

My counterpoint to this is there have always been a ton of bad drivers. The most important skill to have is to be a predictable driver. Bad drivers will always exist. You cannot change them, and the great number of them is not likely to drop to 0 overnight.

The best thing you can do is drive in a way that even bad drivers expect. Be a speed-limited, blinker-indicating, cautious log in the river. They will parkour around you and you will be fine. Greatly limit your reactions to things. If you freak out, others around you will freak out. Bad drivers cause okay drivers to perform worse. Herd mentality.

(Obvious disclaimer: The best defensive driving in the world won't prevent all accidents)

My understanding is currently they do not vet user age. How could they? There are a few billion users. Even among different ethnicities people look younger or older depending on the society they group in. However, I was looking at this from a "confidence score" perspective. If it's an open secret that employees do nothing, throw AI at it to flat likely-to-be-underage users. Then throw what portion of employees you can at that effort, rather than vetting ~2 billion users.

Most of the advertisers and influencers I know personally are children.

Social media was ruined a long time ago. Even here it's a disparate, incomplete conversation. I commented knowing the response I'd get from HN about false positives with AI determining user age.

Time to sue Facebook for the market value of every underage user, and the continuing harm of having a public presence from that age forward. Every job lost, every lost opportunity.

Yes. True.

I think if you use someone's photo for Facebook, they're likely to already be on the site.

You could also have the AI look at the photos that are uploaded over time and if they look very different from the person who supposedly created the account...and comments from others are not questioning why these are a different person.

Right now they don't have people vetting. Throw AI at it for non-perfect but better results. As long as it's not a gov site I'm okay with this. Got off Facebook in 2005 when the first of my relatives became racists.

Presently they don't have the staff or the willingness to vet that all users are above the age of 13.

Nope. That's why I'm not on Facebook, which presently asks for ID when it deems necessary.

See the problem here is I don't view social media as a loss, at least on the level of Facebook.

So I'm wondering why we don't use AI to determine the likely age of a user from the photos they upload. Obviously there will be false-positives but then you can just ask them to upload a picture of their state ID/drivers license.

(I don't participate in social media, beyond HN...)

To me this would seem like a great idea. A couple hundred thousand folks who look 13 but are legally 26 who jump through some extra hoops, but several million children who cannot create a Facebook account.

The risk to reward seems awesome.

This makes me very happy :)

In theory, one can memorize a great deal.. I do agree with you, though.