kuboble

You can still buy whole buckets of generic lego pieces if you wish.

Duplo is also mostly generic parts in sets.

My kids like doing themed Legos as they look cool, but for creative building we do Duplos - even though kids are "too old" for duplo technically. Duplo allows us to build bigger structures. Like https://imgur.com/a/niEhA8E

how do we know they aren't making 9800$ profit on inference from 10k$ tokens at api pricing?

make the proposition of learning to code right now uneconomical

I disagree with this part. The bar of skill you need to write complex now is much lower now, so a new person might very quickly learn just enough to be able to build cool products.

Maybe not operating systems, but useful web apps, browser plugins, productivity tools, programs that solve business needs outside of IT, ETC.

Edit; but if you mean learning to code with the purpose of finding a job as a programmer then I'm more willing to agree.

I think in order to be a productive programmer with Claude you still need to understand what you are doing.

Learning to code is a must. You need to know what is possible or easy to know how much you can ask.

What will be less important is to keep your sword sharp.

I notice that if code less for some time I do more one-off errors, copy/paste mistakes etc. With llm keeping the coding skill warm is less important but I cannot imagine doing my work - even with Fable if I didn't know how to do it without him

That has been my experience at the beginning, but not anymore.

I have developed some intuition of how large tasks I can give it so that it will complete them well, probably erring on the conservative side.

I am using it daily for all my code writing and honestly don't remember the last time I had the feeling that I had to spend a lot of work to get the last few % done.

Well, it's not that obvious.

Some jobs should run every day at 8 am (e.g. torn on the temporary speed limit on front of the school), vs tasks that should actually run every 24h (e.g. feed the bacteria in exact time intervals)

I really wonder about the methodology. The article didn't mention it.

Did they get several cities to participate?

No.

But the other way around sometimes yes.

I don't believe it to be ai generated voice. It's too good.

Or if it is - why e.g. automated voices reading nyt articles are so bad?

I have recently worked on the same kind of similar quiz for German.

However I have some other ideas and my quiz isn't "science based"

- in my quiz there are only "yes / no answers" This way you don't spend eternity reading descriptions of the word "apple". It also means I can estimate separately my passive and active vocabulary.

The OP missing "I don't know button" which will overestimate any result by 25% percent.

- I'm adjusting dynamically how many questions to ask in each bucket.

the goal of my quiz is to estimate a number of German words an English speaking learner has learned.

So I have curated vocabulary to remove "free words" like rare compounds of common words and other rare words which satisfy "any European knows this word without learning".

The final vocabulary used in a quiz is approx 8k words only

https://wortschi.de/quiz

Is it a prompt or accessibility description for screen readers?

Maybe I misread the comment then.

I read it as a models performance being random and observed differences in the opinions are the results of the overinterpretation of the random outcomes.

I think however that some people seem to be always lucky which indicates that it is not random but rather some fixed differences between people and their environments.

No it's not.

It seems that people have different workflows or repos, or memories or prompts or expectations.

I think it's Your mileage may vary.

Few of the best sessions I have ever had with claude went into 700-800k territory.

I frequently reach 400-600k without visible (to me) signs of quality regression.

I think there are at least few question marks.

One being that extrapolating from like 3 data points is hardly science. All trends break at some point.

The other is that the measures to prevent distillation of their models (if it was a secret sauce of Chinese models) could work if nobody is allowed to use them.

Corpus seems to fit and add is missing from the list

Being in Switzerland it looks to me like this is a really tough referendum.

Both sides have very good arguments and from the side it looks like either way the Switzerland has to give up some asoects of its high quality of life.

If the initiative succeeds, Switzerland will get a large hit from the cancelation of a lot of bilateral agreements with the EU.

If the population exceeds 10M then the current rail and road infrastructure will not handle it well.

I have already been on a train which refused to move due overload. And it would only depart if enough people have disembarked. The autobahn are already having hours long traffic jams at peak hours and with extra million people it will multiply.

And it's almost impossible to significantly improve the throughput of rail and autobahn without extreme projects.

It looks like a lose / lose situation is a sense and a people are going to decide which hit to take.

Claude never complains.

In my experience the text for the Claude has only one requirement - the intent and meaning must be there.

The text for Claude doesn't need structure. Doesn't need style. Doesn't need formatting. Doesn't need deeper thought. The only important thing is that it includes somewhere somehow the relevant bits of information.

The quality of prose I throw at him is below what I would show to any other human. I just turn on my microphone, keep dictating whatever comes to my mind and I think might be relevant. After this is done I may or may not ask Claude to rephrase what I wrote before keeping it as memory.

On the other hand people judge you for what and HOW you type. They complain about it.

It's in my experience that people will generally judge a programmer much more for the quality of his outputs than the number of them. So if your target are other humans - it's better to have no docs than bad docs. For claude it's the other way around.

I think the point is that at small scale a single accident poses a risk of ruin to your small operations.

Them you are free to not install them? Why ban them outright?

I'm using https://github.com/cjpais/Handy whichseems to be doing exactly what this app does, and has a very similar background story (author couldn't type die to injury).

Being a domain expert has been more valuable than being an excellent software developer before the ai.

In 2018 I witnessed one guy with no prior coding experience who built a tool that after a month of coding was making very decent money (more than me), just because he was aware of a particular niche.

He showed me parts of his code and it was as bad as my first program, but his was solving a real life problem.

Your observation could mean that overdiagnosis is a thing around you.

I have a close family member with actual attention deficit and extreme hyperactivity.

You cannot mistake it with normal bored kid.

It's like his brain works at 100% intensity all the time.

When you walk with him through the town, he has to touch every door knob, climb every tree, look into every single car, peep into every hole. In a room he finds new object so interesting that he absolutely has to investigate it, for like 30 seconds and then finds another thing and another and another. 24/7. You see his body is tired, almost falling asleep, but his brain won't stop, he keeps finding new distractions he can't ignore.

You would never mistake it for just a regular bored kid.

I was so tired of all those approvals that I switched to Yolo mode exclusively.

Claude works in his own separate vm with root access, git remote set to my local copies of repository no github access etc.

I think he could still hurt me if he really wanted, but most scary stories I heard were about LLM making really bad judgements rather than actively trying to break out and do harm.

?

It got them to be world's greatest power for a century.

My kindle from 2012 used to have ads you needed to pay for to get rid of. It was sold as separate product with or without ads at a time. I had one with ads.

I keep it offline in airplane mode permanently from 2016 and haven't seen a single ad in a long long time.

I can and will still use mine to read files.

What is discontinued is integration with Amazon account. Which seems fair to me to be fair.

To be the advocate of the devil here:

A lot of people believe that spam issue would be largely solved if each email costed 0.001$

Lichtenstein is so deeply integrated with Switzerland and for many purposes is a little more independent swiss canton that it doesn't hold as separate example.

Swiss culture of democracy in unique in the world and is unique among 100% white countries. It's bread among others from its unique geography and history.

Swiss are genetically indistinguishable from French, German, Italian or Austrian neighbors.

When did culture developed, other 100% white countries (which many didn't exist in their modern shape) like Slavic states, Nordic states, Germanic states, French, Spanish, Portugal, Italian, Greek, Balkan etc. All of them have not developed anything like Switzerland.

Credit goes to very specific swiss culture, and the claim that it is what you get of you put white people together... just doesn't hold.

And the swiss neutrality was just a pragmatic decision at a time and today the concept of neutrality is often being questioned in Switzerland while the democratic system is not.

The system is unique to Switzerland.

But it's not the consequence of the skin color or the neutrality. All European countries have been practically 100% white until not long ago with vastly different outcomes.

I think this focuses on the wrong issues.

Tin my observation the main advantage of swiss system is the institution of the referendum.

It means that every major decision is decided by the people. The elected government decides on the 99.9% other issues.

The consequence of this system is that absolute majority of public discourse focuses on the issues and problems and not party affiliation.

So e.g. the most consequential election isn't the one where you have to choose the guy who will make everything great again, but e.g. the referendum if the country should spend billions over the next decades on the new tunnel under alps instead of other infrastructure projects.