Posts by b.trafficlight

OpenAI in throes of executive exodus as three walk at once

People seem to forget that all of those expenses are for training AI models. Right now all companies are still trying to figure out the limits of this tech. It's expensive, yes, but the potential may be still worth it. These billions may seem like scary numbers for normal people but it's not that special considering multi-billion purchases IT giants do every once and then.

One you figure out by trial and error and a lot of research how to build a good AI, I am sure that maintaining its knowledge will be much cheaper and the first "land grabbing" companies to get there will earn it back and more.

Amazon congratulates itself for AI code that mostly works

Re: From the headline

I think it refers to this:

> Amazon Q Developer has the highest reported code acceptance rates in the industry for assistants that perform multi-line code suggestions – with BT Group reporting they accepted 37 percent of Q's code suggestions and National Australia Bank reporting a 50 percent acceptance rate

It's unclear how they calculate this and how these numbers compare to other code assistants though.

AI stole my job and my work, and the boss didn't know – or care

Re: AI is merely Bullshit Generation

Thanks for the link! If this tech started by mimic humans starting to bullshit, it may not be so bad. Gives us time to adjust and prepare in case the tech goes further. https://www.theguardian.com/science/2021/jul/17/martin-turpin-bullshitting-is-human-nature-in-its-honest-and-naked-form

American interest in electric vehicles short circuits for first time in four years

It's the economy plus software

They are running out of people who can buy a car for $50-70k+ USD. Upper middle class with families who wanted one already got Tesla, Rivian, BMW or similar. The rest of the schmucks are getting Model 3, Ioniq 5-6 and stuff for $40k+ but it's not very practical car for a family with children.

Aside from infrastructure, the article should have pointed out that 1) there is not enough lithium to convert all cars to EV and 2) today's EVs suck because car manufacturers suck at developing software. "Car + computer = computer" (Alan Cooper, "The inmates are running the asylum"). Tesla is the only one who was built with that focused (it also sucks due to the whole spying and i-want-to-suck-like-Apple attitude). Other car makers are just begining to figure it out. I mean, they don't even develop their ICE infotainment systems (which suck beyond anything). It will take them half a decade or more to figure out software, security etc and probably with the shove from the regulators in the process.

Study finds 268% higher failure rates for Agile software projects

There is so much wrong in this article it feels like trolling

Like other people commented above - agile is meant to address a problem of building the wrong thing and not realizing it until a lot of money and time has been spent.

"Within budget and on time" as a success criteria is a bit absurd because it implies that you know what you are building upfront to a great detail. I'd argue only a minority of IT projects need and can be done that way. That's when you have other methodologies designed to optimize for end-to-end predictability of estimations and costs at the cost of the speed of development and pivoting (used in aerospace in other systems where cost of defects and changes are high).

How quickly people forget the times when companies would build software for months before releasing it and realizing all the things they did were unnecessary, wrong or no longer relevant.

Also, as someone who works at a big "agile" company I estimate that 90% of all teams which claim that they follow Agile are doing the "Cargo cult" of Agile i.e. standups, sprints and other motions without actually following the spirit. How many of those teams are cross-functional? How many of them have customer or customer representative on them? How many of them communicate and collaborate on their work? I don't think there would be many of those.

So yeah, pick a non-relevant success criteria, apply it to a catchy methodology which nobody actually follows and then claim it fails to achieve that criteria. What is it if not trolling?

Is it time to tip open source developers? Here's one way to do it

The formula which gives larger shares to the top of the tree and less to the leaves doesn't make sense. It would skew to higher level libraries or frameworks and their popularity and undervalue some foundational components everyone depends on.

Russian developers blocked from contributing to FOSS tools

Typical Russian whining

"Oh, poor me, evil / stupid Western companies don't let me contribute!"

I felt sorry for Amelkin for a second - maybe he is indeed a Russian who opposes war etc? Nope. His LinkedIn profile since the full-scale invasion has only a few re-shares of "what-about" articles and other Russians whining about "we don't like that people on the West suddenly discriminate and hate us". No condemnation of the war, no support of Ukraine. Nothing.

It looks like "guilty by association" works pretty well here and Alexander experiencing mild "canceling" does not like it. Public shaming sucks.

I like the approach when Russians are evaluated based on their usefulness to the war. Do they contribute anything which helps Ukraine win? Do they contribute anything helping Russia?

In this case I would argue that Amelkin contributes more to Russian side than Ukrainian side. His contributions to open source do not outweigh him working for a Russian company whose taxes finance the war. Him being oblivious to this only shows how self-centered and disconnected Russians are. Even those who work in IT and have access to the Western media and other alternatives to Russian propaganda.

Apple bags patent for folding phone that closes as it's dropped

It just shows how far patent system has deviated from its roots. It was supposed to protect inventors who invested a lot of time and money into developing something, so that others could not copy that invention and benefit from it. And now we get parents like this, where Apple claims merely an idea of an auto-closable screen. "oh yeah, maybe we'll use a spring, or magnets or something". It is not an invention. It didn't require years of research and prototyping, only 30 mins of engineers brainstorming with the patent lawyer. Sad and pathetic system which encourages such practice...