amoss

It would seem strange to represent doubt by that squiggle though when clearly it is represented by an x in a blue circle.

How would you be offended if somebody referred to you by "they"?

I have not read the paper beyond the abstract so this is something of a throwaway comment.

The setup of the study raises questions about construct validity. The main purpose of these coding assistants is to write new code - i.e. code with a function that is not directly expressed in their training set. The choice of such a simple (and traditional teaching task) for the study means that the function will be heavily represented in the training set. Furthermore, correct approaches to the problem will be represented more heavily than incorrect approaches.

As I said I did not read beyond the abstract so I would be happy to be corrected if the authors have accounted for this in their study.

It is reminiscent of the male behaviour in Universe 25 - this description reminded me of "the beautiful ones": https://www.the-scientist.com/foundations/universe-25-1968-1...

If we flip the question around it seems more reasonable: would your IQ test score be accurate if you did not train beforehand on practice IQ tests?

I would presume not - most tests are timed, and if you are spending time on first-time only tasks in understanding the problem then the result is inaccurate. If you train out those first-time tasks so that you are repeatably using the time budget in the test to solve problems then you should reach some kind of steady state and produce repeatable and more accurate test scores.

My take is that the repeatable scores measuring your steady state in the task would be more accurate than the untrained scores with an unknown amount of initialization time within each problem. I would make a similar claim to naasking below that this could account for some of the Flynn effect.

I'm currently taking a sabbatical for a year. My son is still young enough that he finishes school in the early afternoon and we get to spend most of the day together. It is bliss. The projects that I'm working on still get time in the morning, and in the evening.

Thanks for writing this up and sharing. The rawness of the experience comes through very clearly, and while it must have been difficult to write it reads very well. Hopefully this can help other people in the same situation. Your level of persistence in getting to the root cause is phenomonal.

Given the limited release of access under a license - is there any decision on whether a model like this would be protected by copyright if it leaked?

The output of an algorithm is not a human authored creative work, which given the recent decision in the book case would seem to weigh against it.

Do you think that a real, ancient, stoic philospher would have been able to give a reasonable answer to that question?

Love it, results are legit.

> How should we deal with the inevitable threat of death?

  Friend, death is an inevitable part of life that we must accept. We should
  strive to live our lives in a way that honors the gift of life and prepares us 
  for death. As Seneca wrote, “We must get rid of this craving for life, and learn 
  that it makes no matter when your suffering comes, because at some time you are 
  bound to suffer” (1). We should focus on living a noble life, not just living a 
  long one (2). We should strive to practice kindness, seek out joy, and live 
  without fear (3). We should also practice empathy and understanding towards 
  those around us, so that we build relationships and make a positive impact on 
  the world (4). And finally, we should strive to live a life free from deceit and 
  harm towards others (5). In this way, we can approach death with peace and 
  acceptance.

Takes me right back to studying philosophy in college.

I hope so - I've had chronic insomnia my whole life (and a fair mix of night terrors and sleep walking when younger). Without melatonin supplements I would not have any kind of regular sleep schedule at all. Although on the bright side, screens don't make it any worse.

This makes sense for Apple - they are aiming at the premium end of the market. Having a process advantage over the competition will translate into a performance / battery win. After all, this is what Intel did for decades while they could.

There is a trade-off between the three kinds of improvements. It is possible to pick some points (for example) that give a less than maximum power reduction as well as a less than maximum speed improvment. I think the comparison would be the Apple Silicon that they released is larger than a (hypothetical) alternative that only improved performance but kept the other two constant.

So it is 10 years before GNU and 15 years before Linux... It is also in the era that Microsoft are convincing people that software should have copyright for the first time.

Create the open source movement in a time when there is no dominant, entrenched, competitor. Try out a timeline where an OSS operating system arrives before its commercial competitors in the late 70s.