@mloxton@med-mastodon.com
@mloxton@med-mastodon.com avatar

mloxton

@mloxton@med-mastodon.com

Healthcare improvement and strategic foresight analyst, professional trainer in qualitative research methods. Organizational behavior researcher. Board member at Blue Faery Liver Cancer. #Writer of stuff - #AmWriting

Read my #fiction titles on Amazon "The Screw Turns" at https://www.amazon.com/stores/author/B09P5LLNZ1
See my Kindle Vela stories at https://www.amazon.com/kindle-vella/story/B0BLXHVVSY
... and my children's stories at https://www.amazon.com/kindle-vella/story/B0BLT56MG9

This profile is from a federated server and may be incomplete. Browse more on the original instance.

dancinyogi, (edited ) to random
@dancinyogi@mastodon.sdf.org avatar

Physically, do you feel -

#polls

Please boost for a wider audience 😊

mloxton,
@mloxton@med-mastodon.com avatar

@dancinyogi
Given that most people my age are on several prescription medicines, I think I feel younger than the average person my age

futurebird, to random
@futurebird@sauropods.win avatar

The US Military budget is literally the "I am not good with economy"/"spend less on candles" meme.

No changes needed.

mloxton,
@mloxton@med-mastodon.com avatar

@smellsofbikes @futurebird
This is the bit that really pisses me off about tax time. The IRS asks me for information that (a) it already has, and (b) is provided by the employer, and over which I have no control.

Why are they wasting my time with all this parody?

mloxton,
@mloxton@med-mastodon.com avatar

@tutwilly @smellsofbikes @futurebird
Yup, and politicians LOVE to reward and punish using the tax code, so proliferate complexity and making tax software vendors and consultants seem useful. Inuit et al encourage that, but politicians need little encouragement.

Teri_Kanefield, to random

deleted_by_author

    •
    mloxton,
    @mloxton@med-mastodon.com avatar

    @JessTheUnstill @Teri_Kanefield
    Can you expand on what you mean by "rig the election"

    Are you talking voter suppression, gerrymandering boundaries, flushing people from voter rolls, reducing poll stations in unfavorable precincts, issuing disinformation, ...?
    Or
    Are you talking about changing ballots, falsifying results, hacking voting machines, ...?

    StillIRise1963, to random
    @StillIRise1963@mastodon.world avatar

    Why so lazy as to use chatGPT?

    mloxton,
    @mloxton@med-mastodon.com avatar

    @robertnorlyn @StillIRise1963
    It is wrong about facts because it has no understanding of the world - it just predicts the probability of the next word in a chain of words. It is also generative, so it fills the sentence rather than retrieve it from the data it learned from

    mloxton,
    @mloxton@med-mastodon.com avatar

    @jcrabapple @robertnorlyn @StillIRise1963
    Good plan. None of the LLMs at this point have any way to validate declarative facts, so it would be extraordinarily unwise to ask it to evaluate any facts

    mloxton, to random
    @mloxton@med-mastodon.com avatar

    Not sure I would trust Elon to be putting non-removable hardware in people's brains

    https://www.businessinsider.com/elon-musk-neuralink-fda-approval-put-brain-implants-humans-2023-5

    mloxton, to random
    @mloxton@med-mastodon.com avatar

    I wrote a parts and stock-control program in Algol on a Burroughs B7800, and it would under certain circumstances flash "Don't Panic" in friendly green letters in the status bar

    #TowelDay

    eniko, to random
    @eniko@peoplemaking.games avatar

    i've never really scripted this much story for one of our games before and it's a very interesting experience. particularly the amount of effort required to get around "we can't show that because we don't have the assets or the resources to create them"

    mloxton,
    @mloxton@med-mastodon.com avatar

    @eniko
    Is that much different from the normal boundaries within fiction writing that the plot elements have to make sense, be consistent, and stay within the bounds of believability?

    Like when I make a hospital exec die from hitting his head on a stainless steel toilet bowl, I can give him a brain bleed, but his head can't explode, and he can't turn into a spider and scuttle away.

    mloxton,
    @mloxton@med-mastodon.com avatar

    @eniko
    Well that's a bummer.
    Your plot universe has some interesting obstacles to write around then :)

    mloxton, to random
    @mloxton@med-mastodon.com avatar

    Reason number #3 why we can't have nice things - investors behave like stupid sociopaths who will eagerly set fire to the house to get to a few bucks under the sofa

    Https://newrepublic.com/article/172947/80-percent-shell-shareholders-vote-let-world-burn

    mloxton,
    @mloxton@med-mastodon.com avatar

    @LouisIngenthron
    At this point, the market is largely an engine of stupid

    mloxton, to random
    @mloxton@med-mastodon.com avatar

    META gets fined over $1bn dollars, and Wall St. shrugs.
    This is where we are now.

    In the past, a firm being fined a billion bucks for violating laws would have tanked the stock, and led to resignation of the CEO and overhaul of the board. Now it gets a few days in the news, the stock price climbs, and management shrug it off.

    jasongorman, to random
    @jasongorman@mastodon.cloud avatar

    I've witnessed many "10x developers" actually slowing their teams down, doing what we call "anti-work" - work that creates more work for everyone else (e.g., that massive 5-day refactoring that causes a merge train wreck).

    mloxton,
    @mloxton@med-mastodon.com avatar

    @gdinwiddie @jasongorman
    Yah. My experience is that they leave a trail of inadequately documented changes that create a support and maintenance nightmare that costs 5x the savings that were theoretically gained by having a 10x person.
    They are also often the cause of later prio 1 system down situations

    markwyner, to ai
    @markwyner@mas.to avatar

    All of these creative professionals slipping into AI systems with far too much confidence, hoping for an edge. I wonder how much money AI makers are spending on social-media influencers to post those “don’t get left behind” posts.

    Designers, UXers, and all creatives. I’m getting paid by no one to tell you that you’re talented and capable without AI. Your career doesn’t depend on using it. It’s fun but it’s not our future.

    #AI #ArtificialIntelligence #ChatGPT #Midjourney #UX #Design

    mloxton,
    @mloxton@med-mastodon.com avatar

    @markwyner
    Sure, but I am not going to scroll through stock images or try to draw my own artwork for my #shortstories when AI can do that for me, and having AI summarize the stories for the log line saves me a bunch of effort.

    It doesn't give me an edge, but it saves me bother and money

    mloxton,
    @mloxton@med-mastodon.com avatar

    @ronanmcd @markwyner
    My dude, AI is coming for 80% of the tasks we do, and some entire jobs. No matter what industry you are in, no matter what your job, a big chunk of the tasks that make up that job are going to be done better, cheaper, and quicker by an AI.

    Refusing to use it won't make your life better or your job safer

    mloxton,
    @mloxton@med-mastodon.com avatar

    @markwyner
    There is a very long distance between the true fact that AI used artworks to learn, and saying that the results are therefore theft. If you argue that the AI doesn't create, and hold that modification is not generating new works, then almost no human art would be valid either since all of it is in some sense derivative of prior artworks, and all learning is theft..

    I will continue to use AI art for my images, and use AI to suggest loglines for my stories

    mloxton,
    @mloxton@med-mastodon.com avatar

    @ronanmcd @markwyner
    Thanks for saying so :)

    What does worry me is that very few politicians are actually doing anything constructive about preparing the country for this future, and most of them are wasting scarce time on poppycock and pretense.

    falcon, to random
    @falcon@mastodon.falconk.rocks avatar

    I can't believe I need to say this. If I notice your resume is written by an LLM like ChatGPT - and you can bet I likely will - I will be forced to assume you take the same kind of care with your work.

    Asking me to read or review LLM output is disrespectful of both my time and intelligence. Want a job? Tell me, yourself, how you can contribute to my team.

    mloxton,
    @mloxton@med-mastodon.com avatar

    @falcon
    How would you notice and why in the day of electronic recruitment systems would you ever be looking at the original resume?

    If you are indeed still picking through actual resumes, why do you get upset if someone is using a tool to create it? Are you upset that they used Word rather than a fountain pen? Are you upset if they used your site's resume builder app, or imported directly from LinkedIn?

    Seems kinda unrealistic tbh

    mloxton,
    @mloxton@med-mastodon.com avatar

    @kuhn @falcon
    No, and he would likely be getting far more false positives than detecting actual use of an LLM.

    mloxton,
    @mloxton@med-mastodon.com avatar

    @falcon @kuhn
    Again, you will get a ton of false positives and thereby be rejecting people who have not used an LLM

    mloxton,
    @mloxton@med-mastodon.com avatar

    @falcon @kuhn
    And one last time ... regardless of whether your filtering is automated (bad), done with a tool (worse), or done through your own unaided perception (worst), you WILL be rejecting people who did not use any sort of AI/LLM.

    It is also going to be biased against people that aren't "like you".

    Please don't filter job applicants on ANYTHING not directly related to the requirements of the job itself

    mloxton,
    @mloxton@med-mastodon.com avatar

    @falcon @kuhn
    Qualitative researcher with training in AI here.
    Yes, algos are typically less biased than people.

    People are irretrievably and profoundly biased in dozens of ways, and never more biased than when they strongly believe they aren't.

    Take an Implicit Attitude Test and you will see what I mean.

    mloxton,
    @mloxton@med-mastodon.com avatar

    @falcon
    Nobody is saying that algos are unbiased.

    They are just typically going to be LESS biased than people when it comes to things like assessing resumes.

    In quality improvement, algos and engineering reduces bias and error, and the big challenge is discovering latent or subtle bias caused primarily by using human examples as training data.

    mloxton,
    @mloxton@med-mastodon.com avatar

    @falcon
    Please stop with the extreme strawman stuff.

    Nobody said "all tasks"

    We have over a century of engineering human bias and error out of processes, and there is a whole branch of science dedicated to describing the dozens of inbuilt human biases.

    We also have over 80 years of science around recruitment and job matching.

    There is zero doubt that engineering humans out of many areas of decisionmaking reduces bias and error

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • mdbf
  • everett
  • osvaldo12
  • magazineikmin
  • thenastyranch
  • rosin
  • normalnudes
  • Youngstown
  • Durango
  • slotface
  • ngwrru68w68
  • kavyap
  • DreamBathrooms
  • tester
  • InstantRegret
  • ethstaker
  • GTA5RPClips
  • tacticalgear
  • Leos
  • anitta
  • modclub
  • khanakhh
  • cubers
  • cisconetworking
  • provamag3
  • megavids
  • lostlight
  • All magazines
    •