80% full. It drastically helps extend battery life

In all fairness, in the early days of Google Assistant it really was useful. It actually worked. Somehow in the last 5 years it plummeted. As in it stunningly and noticeably kept getting worse year after year.

The only thing close to a decision that LLMs make is

That’s not true. An “if statement” is literally a decision tree.

The only reason they answer questions is because in the training data they’ve been provided

This is technically true for something like GPT-1. But it hasn’t been true for the models trained in the last few years.

it knows from its training data that sometimes accusations are followed by language that we interpret as an apology, and sometimes by language that we interpret as pushing back. It regurgitates these apologies without understanding anything, which is why they seem incredibly insincere

It has a large amount of system prompts that alter default behaviour in certain situations. Such as not giving the answer on how to make a bomb. I’m fairly certain there are catches in place to not be overly apologetic to minimize any reputation harm and to reduce potential “liability” issues.

And in that scenario, yes I’m being gaslite because a human told it to.

There is no thinking

Partially agree. There’s no “thinking” in sentient or sapient sense. But there is thinking in the academic/literal definition sense.

There are no decisions

Absolutely false. The entire neural network is billions upon billions of decision trees.

The more we anthropomorphize these statistical text generators, ascribing thoughts and feelings and decision making to them, the less we collectively understand what they are

I promise you I know very well what LLMs and other AI systems are. They aren’t alive, they do not have human or sapient level of intelligence, and they don’t feel. I’ve actually worked in the AI field for a decade. I’ve trained countless models. I’m quite familiar with them.

But “gaslighting” is a perfectly fine description of what I explained. The initial conditions were the same and the end result (me knowing the truth and getting irritated about it) were also the same.

This happened to me the other day with Jippity. It outright lied to me:

“You’re absolutely right. Although I don’t have access to the earlier parts of the conversation”.

So it says that I was right in a particular statement, but didn’t actually know what I said. So I said to it, you just lied. It kept saying variations of:

“I didn’t lie intentionally”

“I understand why it seems that way”

“I wasn’t misleading you”

etc

It flat out lied and tried to gaslight me into thinking I was in the wrong for taking that way.

The funny thing about not specifically dealing with misinformation in an LLM is that not trying to account for misinformation will lead to very wild responses in terms of accuracy, and I don’t mean things relating to politics, but things like putting glue into a pizza recipe.

The word “theory” is misused here. It should be “hypothesis”.

A scientific theory is something that’s well established, has a working understanding, and is verified through multiple experiments.

It’s not gamification that’s the issue. That aspect really held my attention and gave me consistency.

It’s the push to a pay-to-win model that made me quit. They made the challenges harder and harder to complete without using boosts, and to use the boosts you had to use gems. And gems were really hard to get unless you bought them with real money. It doesn’t matter if you have a super subscription (or whatever it’s called), you still had to pay to get the gems.

And the prices for the gems were just as predatory and the disgusting mobile gaming industry. Never should there be an option to spend over $20 for in-game consumables, nevermind over $100. It’s sick.

Almost any default text editor on Linux is better than Windows notepad, and many are straight up better than Notepad++

A part of my brain always reads AI as Al (yes, those are two different letters). As in Albert.

So it’s generative Albert. And “Albert is increasingly using more power”.

who have invented their own language about syncing up, achieving alignment, creating action items

My soul contracted in upon itself a little as I read that.

You should read the Children of Time. The second book is something else.

You could extract the private keys from your printer

https://wiki.rossmanngroup.com/wiki/Reverse_Engineering_Bambu_Connect

This is what happened

https://wiki.rossmanngroup.com/wiki/Bambu_Lab_Authorization_Control_System

You might be interested in an episode of Linux Unplugged from a few months ago that goes into Meshtastic quite nicely.

https://linuxunplugged.com/584

There’s also this link with a ton of resources for Meshtastic:

https://github.com/ShakataGaNai/awesome-meshtastic/

You want Meshtastic

Then half of your job is dealing with a shitty company and not being a developer.

Partly correct, but yes.

Half of your job is being wasted on overhead you can’t manage.

Who said I can’t manage it?

How else to describe it than you are half the dev you could be while you blame everyone else. Shit as a director I’m not mixing words.

Director? That actually tracks. Believe it or not, but a good 30% to 40% of a dev’s work is not writing code. You can’t just start screwing 2x4s together and expect a house. There needs to be coordination and collaboration between the devs; and paradoxically, the more devs you have the more time is spent on that collaboration and coordination part. It’s called The Mythical Man Month. Something managers and directors haven’t been able to get into their heads since the 70s.

If you can’t break the problem down part of is on the team but a lot of it is on you.

Break what problem down? What are you on about?

If you want to architect a solution you have to be able to explain it.

Ya, that’s absolutely right. And I’ve yet to meet a PM that can properly explain what they’re looking for. They explain result A, they get result A, then suddenly they actually meant result §∆. Happens every time.

It’s pretty bad having to explain this to coddled engineers learning how the other half of the company works.

I know full well how the other half works. Not a lot of complexity there. 90+% of what’s said in meetings could be an email. I bet you’re one of those directors that insists the devs have daily check-ins (absolute waste of time) and even have some of the other managers and “stakeholders” join to “ask questions”, when anyone who isn’t a dev has zero right and business to be there. Those stand-ups are not for you and you’re just in the way.

Talking as if everyone else doesn’t get it

They don’t. It’s that simple.

why be an asshole about it?

Did you really just spit in the face of an entire group and then cry about someone being mean when they called you out? If you’re a director, holy crap! I pity and fear for the people who work with you.

Why do I get more bs from coders than I do contractors working on my roof?

I have a theory…

My grandfather worked on the Apollo missions so why is C# black fucking magic and suddenly you’re Gandalf?

Your grandfather sounds like he was a smart guy.

It’s not magic. It’s code

To me it’s code and logic. To you it’s magic.

Seriously I am tired of engineers being gate keepers while the other two legs of the stool keep this shit together.

Seriously engineers get your shit together as we are all making a product together

Half my job as a programmer is chasing down the non-devs asking them to explain how they imagined the thing they asked for working, and then trying to find the politest words to say their idea is really bad, all the while trying not to insult their intelligence. The other half is putting out fires that come up all the time because the people who “made the product together” made horrendous decisions about the product design without consulting the devs, or even getting their input. So now we’re saddled with mounting technical debt because of a bunch of morons who were convinced they knew more than the people who teach computers to think.

Seriously, half the things I hear non-devs say make me actually wonder about the “average” level of intelligence of our species.

You don’t work magic.

If it wasn’t, then you’d be able to do the job.

To you, it IS magic.

This is a legit problem with many services now. And some companies are, right now, being sued for this dark pattern practice.

You can sign up with one click, but to cancel they make it impossible. Some companies literally will process an account cancellation only by registered letter. How messed up is that?

Don’t use Spotify for podcasts. They’re actively trying to kill off what we know as podcasting and trying to lock the media into their own service.

Instead, use any app that supports Podcasting 2.0.

https://podcasting2.org/apps

Well, that is true for me in Canada with Teksavvy.