• 0 Posts
  • 24 Comments
Joined 1 year ago
cake
Cake day: July 6th, 2023

help-circle






  • jadegear@lemm.eeto4chan@lemmy.worldAnon lives on the margins.
    link
    fedilink
    arrow-up
    7
    arrow-down
    1
    ·
    edit-2
    10 months ago

    The hard part of app development is identifying and translating all of the nuances of human desires and unanticipated needs into a working application. ChatGPT is on par with junior programmers - it can produce simple programs, help contribute to complex ones, but will struggle (for now) with the complexities of dealing with the major hurdles of software dev.

    Eventually it will reach a point where it can reason about human needs and motivation autonomously (probably stacking multiple specialized LLMs or similar together for each area of reasoning, unless something new comes about) but we’re a ways away from that yet.

    I think the big disruption that ChatGPT will cause near-term is the same as that of generative AI art - the low specialization portion of labor will be replaced, eg stock photo producers and basic CRUD/site apps. For the rest, it will be a tool that gives those that adopt it alongside skill a serious leg up.

    In ten years I think the conversation will be different, but two years to learn means 8 years of good salary and time to adapt to that future. Better than $15/hr* with no healthcare in rural US.


  • jadegear@lemm.eetoProgrammer Humor@lemmy.mlScrum
    link
    fedilink
    English
    arrow-up
    4
    ·
    10 months ago

    Depends. I’ve had plenty of tough calls with management laying out the impossibility of desired schedules only to have the Jira board estimates fudged in their favor, or similar, which puts pressure on the team to deliver on timelines they never would have estimated for themselves.

    Ultimately it’s a question of who’s working by whose estimates.



  • Usually the issue would be that these places cost large sums up front to acquire, and there is inherent risk in lending money or selling something for payment over time.

    The most equitable solution under those circumstances IMO would be a pay-towards-ownership rental model with an agreed stewardship rate for routine maintenance and if they terminate lease early, the accrued funds towards the ownership are disbursed. This allows the “renter” (future owner) the ability to eventually accrue the value of the home without risk of loss of investment, while also allowing the “owner” (steward) to ensure that maintenance can be performed. Would have to work out how to pay for incidental maintenance like a failed water heater or storm damage, but splitting cost across owned percentage may be fair, or based on fault, etc.

    It’s a lot of hassle for something that we should instead fix at the systemic level, but so long as we’re looking at the current system then this ought to do well by both parties and would be accessible for those fortunate/lucky enough to be pulling significant salaries to help those less fortunate.

    Cooperatives are also a good option long-term but I’m thinking in terms of folks that are living hand-to-mouth being able to earn towards a permanent home right away rather than a group of people with enough surplus money to pool for shared home(s). A well-established coop would be a better support network and may be able to grow faster (help more folks) than the alternatives.








  • I’m less concerned with who people learn facts from than with the veracity and balance thereof. If it’s not for you, that’s fine, but there are plenty of people with limited interest in politics that would be willing to listen to a topic that their favorite comedian discusses and reconsider their views on that basis. Gatekeeping politics is how morons keep the greater populace ignorant and disengaged.


  • During meetings, I find it easier to follow the discussion if I’m making notes on post-its or a notepad rather than digitally.

    For longform notes, research etc I prefer to use a wiki program like Obsidian and a mindmap or diagramming tool. I will rarely sketch ideas on paper but being able to rearrange the shapes on digital canvas makes it great for whiteboarding as a software engineer.



  • Disagree with this take in general (growth is worthwhile if only to shift communications platforms in general to open and federated protocols) but I don’t think Lemmy is quite where we need it to be in order to sustain a migration. Finding a good instance is still tough, the idea of federation isn’t easy to grasp for a new user yet, and the UX is still hammering out bugs. (Big thanks to all the devs that already work on Lemmy and all those that shifted over with the Reddit exodus for driving it to new heights so rapidly.)

    An ideal migration from my perspective would have them find instances that cater to their interests and views and would allow easy defederation if undesired. Also, more control for the end user in what communities they see on their feeds when going through discovery (new/hot/etc feeds).

    With better user controls for self moderation and better distribution of users across multiple instances I think we can have our cake and eat it too: growth towards a free world of communications without bogging us down by dealing with the folks/attitudes we find repugnant.