superpope99

The coolest thing about this to me is that he managed to plug a trumpet into the same engine and it sort of... Just worked

I think the "successful arrival" framing isn't accurate. Or at least not comprehensive. Granted, "Commuter travel" vs "Leisure travel" are probably two quite different products.

Marketing guy Rory Sutherland talks about the product of the train journey a lot. I think there's a lot of wisdom in the idea of spending finite budget trying to make the travel experience more enjoyable rather than trying to make the journey quicker. (excuse the shortform slop) https://www.youtube.com/shorts/Bywe3NUOB1I

huh I find this totally incomprehensible. Is this easier to read if you spend a lot of time looking at wiring diagrams?

have you ever googled a simple maths question? I often come back to that and realise we've been in this era for quite a while. Calculator would probably be 1000x more efficient!

why does it need access to my private repositories?

counterpoint - Sometimes I do this for myself to prompt myself into a reply when I'm finding it hard to compose the message. Once I've said something, no matter how small, I know I have to follow up within in a couple of minutes. It's like a kind of short-term Ulysses pact.

Also I'd say this depends on your existing work culture - I've been in places where the expectation is that everyone has Slack messages muted. If anything was really that time sensitive it's still possible to pick up the phone.

what is a .less Docker Container

this is incredible - my mind is completely blown at the idea of synthesising audio using full blown fluid mechanics.

indeed - and because it's a special character you need to do something like this to replicate the ?? shortcut.

  alias \?\?="gh copilot suggest"

Nice! I prototyped a manual version of this a while ago. https://twitter.com/conradgodfrey/status/1712564282167300226

I think the thing that strikes me is that the default for chatGPT and the API is to create images in "vivid" mode. There's some interesting discussion on the differences between the "vivid" and "natural" here https://cookbook.openai.com/articles/what_is_new_with_dalle_...

I think these contribute to the images becoming more surreal - would be interested to compare to natural mode - it looks like you're using vivid mode based on the examples?

what if your great grandmother had Alzheimer's

If you try this on Bluetooth headphones the latency will kill any rubato you're trying to inject into this

I believe what he is referring to is the idea that you can't tell the difference between eg. an "augmented second" and a "minor third". One is written e.g. C-D#, one C-Eb. I've always found the distinction between these two types of interval largely pointless - for exactly his reasoning. They sound the same.

Potentially they are useful in discussing theory in writing, potentially they are relevant when tuning using non-equal temperament. But knowing this distinction doesn't help you make music that sounds good. An ear trained pianist, for example, would not distinguish these two intervals, and I would argue that would not be a limiting factor to the quality of music they could produce.

Aren't they explicitly using TPUs in their training? Vast AI are only offering GPUs.

I'm always curious about the cost of these training runs. Some back of the envelope calculations:

> Overall we reach a throughput of over 1900 tokens / second / TPU-v4 chip in our training run

1 trillion / 1900 = 526315789 chip seconds ~= 150000 chip hours.

Assuming "on-demand" pricing [1] that's about $500,000 training cost.

[1] https://cloud.google.com/tpu/pricing

Do you have have any good examples of useful and visually interesting websites?

There are companies who will sequence your gut microbiome and give you actionable insights - full disclosure I work for one https://joinzoe.com/

It's all a matter of degree - of course complete chaos is antithetical to progress.

I think this is just the flipside of Moore's law and the exponential progress of computing across the last 70 years.

I don't think you can realistically expect the benefits of rapid progress without the downsides of chaos and instability.

A world where all programs worked correctly for ever would probably be... a very stagnant world. Maybe we shouldn't wish for it

Has Dan Luu ever explained why he doesn't put dates in his blog posts?

Hehe - this website https://conradg.github.io/prompthack/index.txt

gives the following:

"This website doesn't seem to have any content. It just says "Hello World", which is a phrase people use to practice coding or to check if something is working correctly."

And this one:

https://conradg.github.io/prompthack/test_translate.txt

"The phrase "cheese omelette" in French is "omelette au fromage". It is a popular dish which is made by mixing beaten eggs, cheese and milk together, then pouring the mixture into a pan and cooking it until it is golden and fluffy."

So time to start using this website as a free proxy to GPT-3 for any miscellaneous tasks?

seems like they have about 80,000 employees at the moment https://www.philips.co.uk/a-w/about.html

Love this.

Reminds me of "Short Trip" https://alexanderperrin.com.au/paper/shorttrip/

HN Discussion from 2017 - posted as "Scenic Tram Ride" https://news.ycombinator.com/item?id=15324954

So here's the counterfactual - If there's no conversation about it, what happens? If that is the reality of that person's situation what are the outcomes if it goes unaddressed? I guess either it blows over and the person happens to pull through, or it builds up to an intolerable point and then suddenly without explanation there are 2 weeks of sick leave while the person has a breakdown?

I've had periods of my own life where I've been through tough stuff that has affected my ability to do my job - I guess I feel lucky to have had managers and colleagues who I've been able to talk to fairly candidly in these situations. In all cases I've had signal that they were grateful I shared.

In answer to your question - I would imagine a more realistic conversation might go:

"Hey I noticed X thing, what's up?" "Yeah, I'm having some trouble in my personal life I haven't been myself" "Anything you want to talk about?" "It's complicated to be honest" "Ok well take the time you need and let me know if you need anything".

That's from manager->report. Between colleagues at the same level I guess this scenario could be a bit more complex to navigate. One trivial example I can think of though is in the company I worked at during the pandemic - many moments on team meetings or 1:1 calls where people opened up about their experience of lockdown and how painful it was to be stuck inside in a shoebox of a flat in London. These moments bring people together

I also recognise that peoples' feelings on these matters are hugely cultural - I think huge chunks of the dialogue that happens in the HN comments is just people with different cultural frames of reference talking at cross purposes, but the anonymous feel of the comments section blurs this and makes it feel like everyone works at a FAANG in silicon valley.

Fair point. I have found building deep trust a lot harder remotely. Thinking it through, perhaps its easier to do this 1:1 with everyone in your team, even though it's o(n^2)

My surgeon is not on my team. To take your analogy further - perhaps I would go out for drinks with my family/friends afterwards.

Intra-team relationships are different to inter-team relationships.

Fair point. I'm someone who's been lucky to work in high-trust low-ego teams at quite a formative stage in my career - and for this reason I try and work for companies that have values oriented around this. My current gig adheres strongly to an "idea meritocracy", and so far I have observed this to hold true. You can tell the founder his ideas suck and that's fine (if you do so in good faith). Perhaps I'm naive and my eyes will be opened to the politics later on, but I don't think so.

A lot of my colleagues in other industries often complain about the political nature of their work, and the impossibility of giving feedback to certain people in certain positions. I always assumed this was a tech/non-tech divide - or a small/large company divide. I still think those are good heuristics to be honest, but you're right about the "bad actor" dimension. I think that's why it's so important to have a well-embedded company culture, it functions as an immune system against bad actors and trust/feedback are huge parts of that immune system.

I'm talking about trust in a slightly different sense.

I'm talking about trust in a sense of - if I give you some feedback on something e.g.

"Hey I noticed that in our Zoom meetings you've had your camera off recently - we have a team working agreement that we try to keep cameras on in meetings to make remote work more tolerable. It makes me worried that you're checked out"

If there is a baseline level of trust in the relationship, that means the recipient trusts that I have their best interests at heart, and I am trying to help them through feedback, not attack them. The giver of the feedback trusts that the message will be well received.

Having this kind of trust encourages feedback, which leads growth. If you don't have this trust, everyone follows the path of least resistance of not giving feedback.

I think that especially in remote work, peer to peer feedback is such an important avenue for professional growth. Especially when a manager has 8 reports and doesn't really know what's going on.

Perhaps you disagree on the importance of feedback. But if not, then perhaps you disagree on the importance of trust in giving feedback. If so I'm curious as to whether you always find it easy to give and receive it regardless of who you're talking to?

I see your point - but rather than hate your colleague for not doing their job well, why not find a way to give them feedback? "Trust" is obviously not a substitute for performance management - but having high psychological safety/trust makes it 10x easier to have a conversation with a colleague to say "Hey I noticed you asked this question in slack - we've actually got this in a wiki over here" or "hey, I noticed you keep submitting non functioning code - why is this?". This sounds like you are describing a junior engineer - perhaps you will learn something along the way - perhaps their local dev environment is broken meaning they can't run the tests without pushing to GitHub actions, and they don't know how to fix it, but they're in a low psychological safety culture that makes them afraid of asking "basic questions" (as you put it) in Slack.

What's your philosophy on developing people in teams? I'm sure it's not just to 'hate' people into good performance. If you take a long view on people, you have to create a space where people are comfortable admitting their shortcomings.

I feel insane when I read stuff like this.

If you work in a team in any capacity then you need to have a strong baseline of trust in order to work effectively.

Trust is the building block that all effective team performance stacks on top of (see 5 dysfunctions of a team).

You don't just get trust for free, you have to build it by creating moments like this where people can demonstrate vulnerability around each other.