netdevnet

What does "evolve replication" mean?

Yes, this is it. The goal of the question is "Do they understand what we and I are doing here? Do they understand what they are asking of me? Can they empathize with me?". And the meaning is simply if they have programming skills.

The article makes it sound like it the usage implies some devaluing of those who are not technical which is not the case (outside tech bro culture)

ReactOs only shares the name with JS library called React

Makes you wonder what future generations will call our English. "Old" is taken and "modern" is always the English of the living

You are saying that after having experienced all the subsequent versions. GPT-2 was fairly good, not impressive but fairly good. People were using for all sorts of stuff for the fun of it. The GPT 3 versions were really impressive and had everyone here super excited

Chat GPT is just GPT version 3.5. OpenAI released many other versions of GPT before that. In fact, Open AI became really popular around the time of the GPT 2 which was a fairly good chat model.

Also, the Transformer architecture was not created by OpenAI so LLMs were a thing way before OpenAI existed :)

such as?

I love it when people post studies just because it agrees with their worldview. For full disclosure, I support the 4 day week but I know to be objective. None of those links contain references to proper RCT with control groups or a variety of office roles so there is really no evidence for it. Trial plus survey is a better term as it carries no scientific connotations.

You are not really going to get your average companies to participate in those kind of studies anyway. Those companies want warm bodies and it has nothing to do with empirical evidence

> I meant productivity overall is constant

That's a huge statement there. Do you have any evidence that this is the case for everyone at every company no matter what?

Could be the 30 most influential companies in Germany.

It's like when FAANG comes up with a new cool shinny tool (MapReduce, React, Tensorflow). Even if not everyone is using it, your will see articles saying "The tech industry now uses [insert shinny tool]" even if most of the companies are running [insert old tool].

Lazy attitude towards proper role management and poor engineering practices. More common in small companies or small teams managing their own service (db and app)

Really all you need is logging and potentially temporary read access to the db if you need some info that you can't derive from the logs.

Totally unrelated but did anyone check the author's CV? He has had multiple intern dev roles even though he had at least 5 years of previous dev experience

And science also has a meaning which has nothing to do with programming. What makes you think that most scientists use computers in this way?

> Autumn hasn't been replaced outside of America though

not sure how is that relevant?

It did in the states. Autumn is no longer a loan word (a loan word has little or no changes). The French word is quite different: automne. As You can see, it is missing 2 letters and the pronounciation is quite different.

If you want to go down the prescriptive path, a more correct word relative to "autumn" is "hærfest". Just like "fall" replaced "autumn", "autumn" replaced "hærfest".

> What makes LLMs difficult is getting access to all the hardware to train them, getting the data, and being able to preprocess that data.

Yes, that's my opinion too. GAOs (Grassroots AI Organisations) are constrained by access to data and the hardware needed to process the data and train the model on it. I look forward to a future where GAOs will crowdsource their computations in the same way many science labs borrow computing power from people around the world.

It was an exaggeration but most of pre-ML AI is just a set of fairly rigid and general rules. They mostly differ in the sense that rules are added by different means (human vs machine) or at different times (during the "training" process or on the go while being used)

ML models are AI too. And you don't even need statistical models to call your system AI, just a set of conditional blocks behind a decision

Surely, if it is AGI it won't do stuff just because it is * told * to same way you as a company don't do exactly what your customer *tells* you to do but what you *think* it *wants or needs*.

So I expect AGI to do stuff we might not understand but further in the future we will look back with enough hindsight and we will understand why it did what it did and we will agree that it was the best action to take at that time. Just like children lacking understanding of why their parents tell them to eat bland veggies or do stuff they don't want to do even though it is good for them in the long-term. I expect this to be the main source of conflicts with AGI in the future.

> A local LLM will make it 100x more useful. Also, it might not even need be "local." If I make it available via the web, I can probably sell access to other scientists and engineers in my field.

That's not legal. Just because you own some files does not mean that you own the IP of the content within.

I speak to them every week and in fact I did ask them this week. It's very insightful. They understand that the phone is on and that it drains the batteries in the same way it did pre-smartphone. The same way an electrical device consumes electricity. The idea of background activity linked to apps (as opposed to general system activity to keep the phone running) they have installed is not something they are aware of. Hence, the first action when it happens is to take it to a technician to "fix" it or buy a new one.

You can desire all you want. Doesn't change the reality. You can't truly have control without understanding. All it means is that users will get some user defaults (and the device will function exactly as it does now) or worse, they will be compelled to change the user defaults to their detriment (and to the benefit of a third party). There's a fundamental information asymmetry. This is the source of the power imbalance and giving more "control" to users (who already struggle to understand the one they already have) won't fix it.

It's not too different from consumer protection laws. Consumers have a limited understanding of their power while companies have a wider understanding. This means that overall companies win and there needs to be a knowledgeable third party constantly watching and advocating over the interests of the consumer.

Techies have this tendency of trying to fix every problem with tech, even when it doesn't work (looking at every problem like a nail that can be hammered). You can't fix this with more tech. A better solution (by no means an optimal one) is to have public free phone tech knowledgeable advocates that people can see in person to talk about their problems and desires regarding their device.

I am not saying what should be only what it is.

I would argue that most don't.

If you ask most senior citizens or simply most users of computing devices to explain what they think about "background services", most won't know what you are talking about. Just because a functionality has been present in a system for a while does not mean that everyone will know about it. Ask people about dual booting, most won't know about it. Ask most people about "SSL", most won't know about it. Ask users about the command line, most won't know about it. And so on. Even the abstract concept of a search engine (separate from the concept of Google Search) is likely to fly over the head of many people. You vastly overestimate your audience. Let's put it this way: most computing devices nowadays are mobile phones and for many people this is their first and main interaction with computation. The modern ubiquity of phones is bigger than that of personal computers by an order of magnitude imo. Personal computers were huge but mobile phones have a level of penetration PCs could never reach.

When you buy a device, you agree to the terms of service. Most manufacturers like Apple have something along the lines of "don't interfere with the proper working of the system". Freezing apps sound like it breaks the ToS.

Buying a device is less like buying a toothbrush and more like buying a tractor. You don't own it inconditially. You own it under certain terms. Those terms give the manufacturers lots of power.

It's up to the developer. Users don't know what background services are

Techies want control and customisation. People want convenience and smooth interop. The most used desktop OS is Windows not Linux. HNers often forget that they live in a tech bubble. Reminds of me when Dropbox was dismissed because anyone could write a little script to replicate the Dropbox functionality.

Not every app needs to be a web app or use internet yet most apps nowadays do. Either you allow ALL apps to have it or you don't. You will think that devs would make apps that don't need background services. It won't happen. What will happen is that every time you open the app, you will be nagged to allow background services to run. The only thing that would stop it is for that feature to not be available at the application level.

My point is that the suggestion is effectively asking for the removal of background services

Are you saying that people don't want the features that background services provide? Features like notifications

Doesn't this "freezing concept" effectively kill the idea of background services?

This is why I highly dislike the term "coding". It implies that the basis of the role is "typing code"