Takumidesh

You kind of get it, it's not really a dictionary, it's more like a set of steps to transform noise that is tinted with your data, into more coherent data. Pass this input through a series of valves that are all open a different amount.

If we set the valves just perfectly, the output will kind of look like what we want it to.

Yes, LLMs are prone to hallucinations, which isn't always actually a bad thing, it's only bad if you are trying to do things that you need 100% accuracy for, like specific math.

I recommend 3blue1browns videos on LLMs for a nice introduction into how they actually work.

I'll just say, it's ok to not know, but saying 'obviously' when you in fact have no clue is a bad look. I think it's a good moment to reflect on how over confident we can be on the internet, especially about incredibly complex topics that cross into multiple disciplines and touch multiple fields.

To answer your question. The model is in fact run entirely locally. But the model doesn't have all of the data. The model is the output of the processed training data, kind of like how a math expression 1 + 2 has more data than its output '3' the resulting model is orders of magnitude smaller.

The model consists of a bunch of variables, like knobs on panel, and the training process is turning the knobs, the knobs themselves are not that big, but they require a lot of information to know where to be turned too.

Not having access to the dataset is ok from a privacy standpoint, even if you don't know how the data was used or where it was obtained from, the important aspect here is that your prompts are not being transmitted anywhere, because the model is being used locally.

In short using the model and training the model are very different tasks.

Edit: additionally, it's actually very very easy to know if a piece of software running on hardware you own, is contacting specific servers. The packet has to leave your computer and your router has to tell it to go somewhere, you can just watch it. I advise you check out a piece of software called Wireshark.

Why do you think it needs an Internet connection? Why are you saying 'obviously'

Well an uppercase ASCII char is a different char than its lowercase counterpart. I would argue that not differentiating between them is an arbitrary rule that doesn't make any sense, and in many cases, is more computationally difficult as it involves more comparisons and string manipulations (converting everything to lower case).

And the result is that you ultimately get files with visually distinct names, that aren't actually treated as distinct, and so there is a disconnect from how we process information and how the computer is doing it.

'A' != 'a', they are just as unequal as 'a' and 'b'

Edit: I would say the use case is exactly the same as programming case sensitivity, characters have meaning and capitalizing them has intent. Casing strategies are immensely prevalent in programming and carry a lot of weight for identifying programmers' intent (properties vs backing fields as an example) similar intent can be shown with file names.

Being able to back up and then encrypt the messages on cold storage for when I may need to go back through an old conversation doesn't negate something like disappearing messages.

It's the best of both worlds, messages go away over time so if you lose your phone / it's compromised, you don't give up the goose, but you also have a nice safe stored version in the off chance you need it.

The danger imo isn't in having the messages at all, it's more about how, when they are just on your phone or whatever, they are generally not locked down.

It also has nothing to do with DEI, which is generally about workplace culture and hiring practices.

It seems a lot of people are confusing DEI for affirmative action.

A company taking excess revenue and reinvesting it isn't profit.

There may be phrases with the word profit in them that include that value, but the general accepted definition is profit is the money that gets distributed to the stakeholders after expenses are covered. This is things like dividends for publicly traded companies, or for private companies, it's just straight up paying out the cash to stakeholders.

So a service company that only pays salaries has 100% profit?

This is splitting hairs and if all the people arguing about this took an actual class in uni a out this they would know that.

Gross profit typically includes cost of goods sold, COGS doesn't have an explicit legal definition, it's up to the business to decide what they include, they can include employee salaries or not, this is called abortion costing, a business which puts salaries, rent utilities, etc, under abortion cost would have a gross profit equal to their net profit.

When dealing with accounting, you can call things whatever you want, net profit isn't something that has a legal definition.

For example, I just decided that my business doesn't follow your definition of profit, and instead defines profit as only money I find in my pockets. There isn't a legal definition of how I need to define profit, so it's just as valid as all the other definitions.

And regardless of all that, I don't understand how anything you said proves me wrong. Profit is net profit, just the same as profit is gross profit, you can put an arbitrary boundary at any point in a financial metric and it makes sense to do so, but it doesn't change what the word profit means. But the claims that 'if you don't profit you have to go in debt' is just silly and only makes since if you cherry pick a very narrow definition of profit that is used as one part of a general financial metric for a business.

A company that has revenue - all expenses = 0 does not need to be in debt, this is also how a non profit will look, 1 million in revenue, 500k in general expenses, 500k reinvestment into the company final result 0 dollars left over. The effective meaning and understanding of profit for practical purposes and lay people (not book keepers within a company that needs more refined and specific metrics) is the amount of money that gets distributed to stakeholders after a company has covered its expenses.

Your block about non profits is exactly my point. A non profit does not pay out the left over money to stakeholders but people who work for a non profit still make money.

Non profit means no profit. Salaries, rent, etc are not paid from profit.

That is fundamentally what profit is, revenue less expenses. By definition, profit is money that does not have anywhere to go in a business, and so gets distributed to stake holders of the company.

I mean, non profits exist. Of course it's not the case for Amazon, but you don't need to profit in order to exist as a company, and people still get to make money.

Yes, but revenue pays for things like salaries, advertising, etc.

The internet as a whole is also different than it was 15+ year ago.

Your typical reddit user now doesn't know what digg is or who Kevin Rose is or the significance of something like 09 F9, sites, people, and events that were big things on the internet are just nothing now.

Hell, tons of users on reddit were literally babies when reddit was young.