Friend participating in some sort of simulated glider tournament trained a neural network to fly one some way (don't ask details). I recall rules were changed to ban such, not because of him.
Using Claude sounds overkill and unfit the same time.
Feels like that there was a World War started on smaller spark than some of those in the OP in a tense world. And this world is tense again, very tense.
> We buy it to be a tool in our toolkit, not the center of our lives.
Somehow this thought evades almost all software providers nowadays. Distracting, and in equal amount, obstructing self promotions and unwanted hints are the norm, which is making life worse, not better. Then why paying them to annoy us? Somehow this very basic thought is not there, not a bit in most of the cases. They want to be the center of our attention. Idiotic. (I do not ask for forgiveness for this strong word. I believe that the costly - and eventually paid by us - marketing teams are dumb copycats senselessly pushing bad practices established elsewhere through time. I do not dare trying to find exceptions, it is hard.)
> I know the reason this happens is because we aren't the real customers of an app. Nor are the creators / partners. The real customers are the shareholders.
Exactly.
I am in an engineering design software developer organization bought by an investor from the founders approaching retirement (they worked 3 decades on this software helping construction engineers designing better homes).
Ever since the lead up to the sell - changes were tuned to lure in investors, for the liking of investors - our organization is focusing on maximising revenue. Fast. That is THE focus. New marketing strategy, sales strategy, licensing strategy changes, reshape organization to have more informed decision making in sales (i.e. collecting and processing much more data on increasing number of contacts). Company meetings are about EBITDA, sales goals vs. actual, streamlining organization. Luncbreak discussions evolve around how to license existing features differently so it would trigger/force up/cross sales.
What is not on the agenda for maximising revenue: features and engineering. We are a "sales oriented organization", says our new CEO prodly - brought in during the sale. Addressing user needs and becoming more popular for the eventual income boost takes longer than the sales cycle of less than 5 years (the investor wants to sell the company in 5 years time). Engineering is in the way, accounting books need to look much much better much sooner for the eventual profit. Only sales tactics work here.
I see ralted pattern elsewhere, in tools I have the misfortune to use (SaaS and other subscription based products). Shameless self-promotions (cross-sale) distact your focus all the time, 'features' good for the assumed 'cutting-edge' image of the organization, privacy offensive practices (data for running sales campaigns), 'offerings' that help you with the ideas they force on you for some sizeable extra cost.
It will not end well. Takes long time to fail, but without valuable features and engineering there will be no value left for the users to buy eventually. No user wants top notch marketing, licensing, and sales strategy for the benefit of the organization.
We may be surprised how illefficient companies are in organizing the creation of sophisticated things (including processes) for themselves, to use (so for the cost center column).
Higher management figures out things to do in strategic level, in brief, and pushes on "soldiers", who kick it through in the least time (cheapest of the cheapest, for the sake of the quarterlies) EXACTLY the way management told it. Because they have to, their job is to make happen the company objectives given, the way it is given. Pushing out crap in the shape of the thing expected.
Larger organiztaion can use these kind of things the most. Even if they don't do that.
I tried it briefly and the practice - argued for strategy for operation actually - to override my working folder seelction and altering to the parent root git folder is a no go.
If it is so simple - that ensures nothing, faked easily - then what's the point wasting efforts on it? Why to complicate things? Why spend time and efforts to do it? And annoy with one more tiny thing on top of the hundrends? Why not just not doing it?
Or, in contrary, when it is very reliable, so it can map a very specific real person to a reliable and true birth date, then f off binding myself to a randome computer account that gives it out to whomever is asking it!
Using Claude sounds overkill and unfit the same time.
reply