swid

The halting problem is unsolvable due to a contradiction, but there is a finite set of programs with given number of symbols; one or more of them share the max non infinite running time, and if you did have the oracle to answer that, you can in fact run the algorithm provided, except for how long this would take in real time.

A different take is that you can write a program that will only terminate if ZF set theory is false, which is impossible to prove using math due to Gödel's incompleteness theorems. We assume this program should run forever, but would not be able to prove it. That program provides some upper bound to the highest BB number we could ever hope to prove with certainty: 748 according to this article from a few years back.

The problem of the post is that an all knowing oracle doesn’t have this problem, and assuming the program has finite length, our being able to prove whether it halts or not does not stop there from being an answer to that question.

https://www.quantamagazine.org/the-busy-beaver-game-illumina...

FWIW, the most recent thing I built that could be considered a DSL is how we define roles where I work. It’s a lot like IAM policies. I wouldn’t want to run arbitrary code when parsing these documents, and I think yaml suits it well.

I wanted to say - I’ve used GPT3 before when a normal thesaurus fails me… I think it’s possible to find a solution with that kind of technology to get more information for the user than a normal dictionary based approach achieves.

I do have issues with how yaml deals with strings, but it’s the best popular format I think I can use, and seems appropriate for developers.

I will probably use YAML for the foreseeable future instead of creating new languages or interfaces. Think like how CircleCI configuration works. Essentially it is a DSL, but implemented in a format that is widely used and easy to write and parse.

This was a problem for us when applying a template to the schedule - templates could have over 500 appointments (so difficult to apply the template in 1 tx), and not being allowed two ranges meant we could test the appointment starts but not the end. The collection issue came up because we had appointments off of users off of orgs. It was impossible to search all the users within an org/tenant the way we set it up. Either search in 1 user or all users in all orgs.

We ditched all of firebase and I’m happy we made that decision early enough.

From [1], with my own editing...

When comparing the difference between now and human performance

> ...[huamns] can achieve closer to 0.7 bits per character . What is in that missing >0.4?

> Well—everything! Everything that the model misses. While just babbling random words was good enough at the beginning, at the end, it needs to be able to reason our way through the most difficult textual scenarios requiring causality or commonsense reasoning... every time that it lacks the theory of mind to compress novel scenes describing the Machiavellian scheming of a dozen individuals at dinner jockeying for power as they talk...

> If we trained a model which reached that loss of <0.7, which could predict text indistinguishable from a human, whether in a dialogue ...how could we say that it doesn’t truly understand everything?

[1] https://www.gwern.net/Scaling-hypothesis

Edit to say - he was making a joke, this will help you understand

Again, it’s not that I disagree with most of your premise, but in fact, as the comments (not just mine) show - people have issues with submitting too early, so it can be a good decision to break that depending on context. Fewer errors. Which is why I brought up how the problem is presented to the candidate. You might be filtering out people based on something less universally accepted and understood than you think.

So perhaps that isn’t the best litmus test? I wonder if it’s written down that should work as part of the assignment.

In the reverse direction, you can try:

A horse rides an astronaut

And you will probably generate an astronaut riding a horse. It’s not a poor description of what we want; our assumptions about how grammar should work aren’t being honored.

> there’s a difference between number of characters and width

Billboards use large letters to fill the space, but normally they would be readable on a business card.

Personally, I think websites should fill up the available space, increasing font size if it makes sense, perhaps using resizing images. So, no - text should not be constrained to a maximum width. If width were a problem, billboards wouldn’t be a thing.

But if I found out and I was working for you, I might question how much you care about making the company successful. You've already said the salary is not competitive, so without equity; you can easily pick up and leave as soon as you find a better role. Maybe reading the comments in this thread will even make you realize this; you could be at a new job in less than a month.

I think I can say this isn't just "your" problem - you've stated the situation is demotivating and you are not giving it all as CTO; passing over "great ideas". Being unmotivated at work is a personal problem if it bothers you. Passing over great ideas or changing jobs is a shareholder problem, but that doesn't include you. It would potentially be something to bring up to shareholders especially if you are prepared to leave over it.

And FSD is still not full. So what limited use case is this robot for? Boston Dynamics doesn’t seem to sell that many of theirs. Maybe Elon wants to send them to space.

The simulation over produced energy from green source. Looking at the chart - looks like the wind power alone was enough to satisfy the grid if available. Given that, I do think it's possible that a mere 5 hours of backup can keep the grid 98% renewable, turning to hydro when solar + wind is not enough. Also solar / wind may be somewhat inversely correlated and north / south Australia also may have different weather patterns.

Given a large enough sampling of the world, I'd imagine it would be easier to find the wind / solar power you want, although politics could potentially make that harder.

https://news.ycombinator.com/item?id=32567994

> During the cooler winter months, more than one-third of particulate pollution comes from wood smoke. On some days, in certain areas, wood smoke can account for up to 90 percent of particulate pollution.

> On cold, calm days, wood smoke can become trapped close to the ground by an inversion layer, typically formed when a layer of warm air acts as a lid over a layer of cooler air. Inversions prevent the air below from rising, which causes pollutants to build up.

https://www.sparetheair.org/about/what-is-spare-the-air

If you think the optimal amount of crime is greater than zero, at some point we are clearly using different applications of the word optimal. One person is talking about the level under the optimal “solution”, while the other is talking about one constraint that still must be balanced against other constraints. The optimal amount of fraud spending is zero, but then we’d be left with a ton of fraud.

https://open.spotify.com/playlist/2Z69DiD4sGs5aHt4OXE3fU?si=...

So there are two ways we cannot see something - if the universe is not old enough for light to reach us yet, or if there is too much new universe being created in between us and the light for the light to catch up to us.

In a far enough future, there will be no other galaxies in the sky as all the matter will be too far away for light to reach us.

I wrote the student vm system for udacity, and I spun up student vms before they needed them, with some last mile loading to finalize the files they need. The student VMs were not using k8s, although a small piece of the infrastructure did.

I worried most about untrusted users working in a complex environment with the ability to harm the experience of other users, and just used GCE.

For me, boot time was < 5 minutes, so if you can predict the next five minutes of demand, you can boot those machines early. If you are wrong you will pay extra or someone will wait extra time, but still less than the full boot time. Generally it takes less than 10 seconds to access a vm with your coursework on in.

People always ask if the brain is a computer, decide it is not, and come to the conclusion that sentience would not work on a computer. I think the universe is the computer / algorithm and the brain is the data being fed into it, which leads me to say consciousness can exist digitally.