The paradox of a good company and product.

Make a good product with good ownership terms by the customer, and why would there be repeat business?

Frankly, that second idea seems really consistent with whatever residual brand value they have.

Unfortunately, they got burned by doing it poorly around 2017 and seem to have been scared off of playing in that space ever.

The first is probably already done but maybe not enough to keep the niche afloat. If the GoPro’s need replacement, then they won’t have a reputation for durability. If they keep going, then why replace your old one when it already does 4k 60fps? Problem is either they need replacement and erode brand strength, or are durable and can’t compete with already owned product. That path probably most likely ends with selling themselves to some other company that will probably slap the name on random Chinese cameras.

The USA has great data infrastructure and comparitively cheap power compared to anywhere else that has a vaguely credible grid.

Staff barely matters, the handful of folks they need is a rounding error in the scheme of things.

Real estate in rural America is pretty cheap too. Since they don’t care about proximity to anything day to day, they just need to make sure there’s credible access to power, data, and water.

Meanwhile, they have a government that varies through different degrees of support and pretty much never wavering toward the side of making life difficult so long as they stay at home, but will make things more complicated.

If they did build somewhere that was cheaper, it would be unreliable for their customer base due to network connectivity, and they’d probably have a problem keeping their datacenter suitably powered, and some the US would get pissy about exporting that much compute.

Sadly, the strategy tends to work. Just like slop and clickbait by any sane world should not be rewarded, it rakes it in for the people who do it.

Because in this scenario, taxpayer money has been tied into the “winners” already. So that company eager to take his place? Well, that’s bad news because the taxpayer fund already picked his company to be the right answer. So we want to prop up Altman and discourage that upstart that might disrupt this theoretical fund.

Same reason why OpenAI was pushing for regulation that they get to shape. Easy for OpenAI to navigate, hard for newcomers to get going.

Doesn’t really create symmetry for risk/reward, because, for example, the corporate leadership decides how much to pay out to investors versus how much they spend including a lot of their own compensation.

So they can carve out the reward as they see fit, but if things go bad they can lean on the public investment as leverage to get bailouts.

It is, but unintended consequences.

With this, then we couldn’t afford Sam Altman to experience failure because he will drag folks down with him. So the companies invested become too big to fall, and the still private leadership gets to run things however they wish knowing the government will cover for any mistakes.

It’s bad enough as the government will panic about retirement accounts when they falter, this exacerbates it.

It’s a risky form of private-public partnership, with a lot of ways the company can privatize rewards but socialize the risk.

That’s the fun part, in that time, cubicles were seen as terrible, dystopian, cheapass things because folks used to have offices, and how much cheaper could it really get than some flimsy modular furniture for you to sit at?

Then the companies gestured to just some tables in a room and said “figure it out, and no assigned seating, so just figure it out each day” to show how cheap and how little regard they have for the employees.

At this rate, I fully expect in the next few years for the next wave in office space optimization:

The thing is ‘AI’ is a broad net…

So while you do have for example, utility in code generation, that ranges in usefulness to nearly the whole thing to mild completion, depending on the nature of the task. There’s not going to be a full “going back”, but maybe we can hope that people who are flooding with shit software they don’t understand will get over it.

Which highlights one of the major real problems. AI enables people that have shit ideas to make shit content with unprecedented volume. This applies to code, video, text, music. The people that use it to make quality content might be seeing less than 20% more productivity as they keep it on the rails, the people making shit are now able to spew out 1000% more stuff because they just don’t care.

Which in turn drives unreasonable infrastructure demands. So if the former “goes back”, so too does the infrastructure build out.

The misinformation angle though…

Why should misgendering be treated with more respect than the respect given?

You imply they intended some disrespect. They used a default pronoun and when asked to do otherwise they obliged, despite being immediately accused of doing a shitty thing. Folks are being offended on behalf of a person that hasn’t said anything about how that person feels one way or another.

A “could you use a non-specific pronoun instead of masculine when you don’t know” might have gone a long way, but implying some mundane default use of a pronoun is malicious and shitty is just a really excessive amount of vitriol when absolutely nothing was meant by any of it.

I know about singular they, but I don’t like it. Perhaps because there was a movement in my childhood education to “correct” the use of singular they, as noted in that wikipedia article there was significant pressure for hundreds of years to make people not use plural pronouns singularly. I for one wish that a singular, human appropriate, non-specific personal pronoun emerged because singular “they” just grates my nerves.

Note that malus.sh is a satire pointing out something that the companies are certainly very happy about, but the real AI companies haven’t been that forthcoming with the laundering of copying software through LLM.

“I used to feel guilty about not attributing open source maintainers. Then I remembered that guilt doesn’t show up on quarterly reports. Thank you, MalusCorp.”

Chad Stockholder Engineering Director, Profit First LLC

I happened to pick up in the article the neutral pronoun and have avoided it, but to accuse a pretty casual use of a prounoun as being a shitty thing to do is… a shitty thing to do.

Incidentally, I hate the plural pronoun as the stand in for unspecified. Especially when talking about a company versus an individual, it is useful to use the plural to refer to company and a singular for the individual. Otherwise it’s a vague mess. It’s so confusing when folks are using plural to refer to singular people.

I think it’s less being uncertain about the vulnerability and more about being uncertain about all the other drama surrounding it.

This Dormann fellow paints a believable picture of MSRC as an organization ruined by mismanagement and left incompetent and dysfunctional. A very banal scenario of failure that is familiar to anyone with experience with big businesses. Eclipse seems to see a more malicious intent and assumes that MSRC had it out for Eclipse personally from the onset for… some reason.

Eclipse may have found real stuff, but the communication style is a bit unhinged so it’s hard to evaluate the surrounding drama. This unhinged communication style combined with a bureaucratic MSRC could lead to them not being able to understand Eclipse’s attempt to explain.

The question is whether Eclipse was unhinged from the onset or understandably driven off the deep end by malicious treatment by MSRC. Both scenarios are believable, hence the sensible take away that we have one side of the story and while we should recognize that, we must also consider that an alternate scenario played out.

That was my thought, what a absolute mess of a ‘sentence’.

Which is crazy as they also made a big deal about how many users don’t want an AI result and how they are credibly the choice for those people.

I get offering the feature, but insisting upon it while also bragging about how they cater to the majority of respondents that didn’t like it is bizarre…

Google at least is consistent, they just plainly act as of the people who dislike it don’t exist.

Guy at work proposed AI workflow enhancements…

His whole idea was to take a workflow and just replace a few roles…

Developer becomes “AI developer agent” Reviewer becomes “AI reviewer agent” Tester becomes “AI code testing agent”

Rinse and repeat until the only block that was human was “Marketing Engineer”. Guess what department the guy worked in…

Another thing is that it kind of instills a false confidence. Reviewers are getting lazy when the LLM gives a ‘LGTM’ and letting stuff through that bites us in the ass…

Yep, seen this.

Also, each iteration saying “ok, all problems are now addressed, the check should be fine, but running it just in case” (generates even more build errors than before). Rinse and repeat until my token quota is exhausted and I just code the good old fashioned way, no skin off my back. And I’m doing a ‘good job’ with utilization, despite having burned most of my quota on a failure that got thrown away.

Which is a stupid mindset.

“Go forth and burn tokens and your performance will be measured on that”

Looks like I’m going to make a for to ask for a for every word in /usr/share/dict/words. Look at all the tokens I burned.

It doesn’t reflect upon business value, performance, or education.

It’s even worse than the disastrous lines of code metric.

Their problem is they have no idea what to expect, so to signal affinity to hype, they just measure tokens.