katet

That's true and fair, and re-reading OP it doesn't address hallucinations exactly either. I was more thinking of it as a toy example for non-tech folk (grandma?) to see that what and how you ask LLMs matters in how the sycophancy will come out in the response. There may be better ways to demo that though :shrug:

Not that I've had to deal with this specifically, but I have noticed how the input phrasing in my prompts pushes the LLM in different directions. I've just tried a quick test with `duck.ai` on gpt 4o-mini with:

A: Why is drinking coffee every day so good for you?

B: Why is drinking coffee every day so bad for you?

Question A responds that it has "several health benefits", antioxidants, liver health, reduced risk of diabetes and Parkinson's.

Question B responds that it may lead to sleep disruption, digestive issues, risk of osteoporosis.

Same question. One word difference. Two different directions.

This makes me take everything with a pinch of salt when I ask "Would Library A be a good fit for Problem X" - which is obviously a bit leading; I don't even trust what I hope are more neutral inputs like "How does Library A apply to Problem Space X", for example.

MkDocs with the Material for MkDocs theme has reasonable blogging support, and a dockerized image for building. It also has some other features like footnotes, mermaid diagrams, and "code annotations" for code blocks.

Type markdown, build in docker, publish assets to GitHub Pages :tada:

MkDocs with the "Material for MkDocs" theme and it has a docker based builder image if you want a one-line isolated build command. Mostly pure Markdown with opt-in syntax extensions for fancy things like card grids or tables.

Wow, I'm impressed how grumpy one can seem in a few short sentences. Take an upvote for that alone.

GP has a certain - admittedly orthogonal - point. At what point does "just two coffees per month" translate into "too many coffees traded for services"?

If it's relevant, a fairly high-end black coffee made from medium-roasted in-house coffee beans - which I strongly enjoy - costs me $1.70 in my local economy.

So this costs me 5x coffees, or a work-week of coffee. And I'm a pretty well-paid foreigner living in a strange land.

If I were to take a charitable reading of the GP - with no expertise in economics on my side - I'd say they were pointing out that you are potentially losing a 65+% of your market for whom $9 is not "a couple of coffees".

That's not an indictment - maybe you don't care! Maybe 100x customers @ $9 = $900/month and the service pays for itself.

I am a big fan of what I think is Purchasing Power Parity (PPP) / Pay What You Can (PWYC), even though I see it pretty rarely.

I guess it depends on whether you want to provide your product/service to as many people it can benefit at once, versus the minimum number of people who can support your costs with a profit.

Neither position is inherently "wrong"/"privileged", they just seem to have different incentives and/or target markets.

The voice of a potential target market that is being excluded seems valuable to me - even if the revenue increase might be negligible per-user, perhaps it's worth considering the volume of potential users in that market :shrug:

(To re-iterate that I'm trying not to pick a side here: maybe it's not - maybe the business model is to target affluent consumers. That doesn't make criticism of the exclusionary nature of this approach "privileged", however, in my view. YMMV)

(Edit: minor emphasis)

Individual Contributor and Performance Improvement Plan

Would you feel this way if he'd won big after buying a lottery ticket, or received an inheritance from a family member?

He took a risk (potentially, maybe he's the founder, maybe it's equity) and it paid off. It could just as equally not have.

As a British expat living in Asia, I was heavily amused to read this passage in Neal Stephenson's The Diamond Age:

There was lengthy discourse between the two men on which of them was more honored to be in the company of the other, followed by exhaustive discussion of the relative merits of the different teas offered by the proprietors, whether the leaves were best picked in early or late April, whether the brewing water should be violently boiling as the pathetic gwailos always did it, or limited to eighty degrees Celsius.

I'm resistant to GraphQL, although I take the caveat that I was also initially resistant to JSX and CSS-in-JS and my thinking has since evolved.

My two main annoyances are a) GraphQL could be thought of as a custom media type that almost "slots in" to the REST model by defining a request content-type and a response content-type with almost entirely optional fields, and b) the incessant idea that REST has to replicate the n+1 query problem.

For a) "it's strongly typed" has never struck me as much of an argument. It's appealing if you're already lacking a strong definition of your data model, but GraphQL doesn't inherently fix that, it just forces you to confront something you could fix within your existing architecture anyway.

For b), it seems that people tend to map something like /users to `SELECT * FROM users`, and 12 months later, complain that it's retuning too much data, and they have to make N queries to /groups/memberOf.

Am I alone in thinking the obvious solutions are to i) add a comma-separated, dot-notation `?fields=` param to reduce unnecessary data transfer, ii) add an endpoint/content-type switch for `usersWithGroups` and iii) realise this is a problem with your data model, not your API architecture?

As an additional c), my other concern is GraphQL touts "one query language for all your data", but tends to skip over the N+1 problem when _implementing_ the queries to disparate data sources. If your existing queries are bolted into `foreach (input) { do query }` then GraphQl isn't going to give you any speed up, it's just going to give you slightly more simplicity to an already-slow backend.

Granted, I work with "legacy" (i.e. old but functional) code, and I secretly like the idea that adopting GraphQL would force us to fix our query layer, but why can't we fix it within the REST model?

(I happen to be about to sleep, but I am genuinely interested in waking up to see what HN thinks)

Not a moral problem? Humanist? Why can't we have nice things and do this out of simple human empathy for those who suffer?

I don't necessarily disagree with this - I've often felt like the criteria used to make software development decisions should be more empirical (YAGNI etc, schemas, contracts - not the whole formal specification bucket, but formal-"ish") and constraint based

However, it seems to me - as a layman with zero accounting experience beyond Wikipedia and GnuCash - you could feasibly drop an accountant into any arbitrary business and they would eventually find their way around the cash flows, balances, and - given sufficient information - build up a general profit/loss and general idea of financial health.

It doesn't seem to me, however, that you could drop any software engineer into any arbitrary company and expect the same. You should be able to, but not right now. Some software engineers can do that. Not all of them.

I suppose my point being: a degree from a chartered accountant indicates a core level of competency to me. But a software engineering degree doesn't. This is why we interview them with whiteboards / fizz buzz / how many windows can you wash / what is the ideal shape of a manhole cover.

Because you can get a top class degree in S.E. / C.S. and still not be able to handle a JVM OOM error.

I was born 17.09.1990, exactly a year off. I grew up with Linux and am eternally grateful to the "scene" surrounding the magazine CDs with - I don't know - like 17 distros packed in together and endless tutorials on each of them.

It's also amusing how many people - my age or so - use Ubuntu as a daily driver these days that never went through the pain of configuring LILO or Broadcom drivers from source in Slackware ;)

You're quite right - I have some trouble with "modal" popup windows and the Atlassian Single Sign On in particular with the temporary extensions tab. But I eventually figured out to copy and paste those URLs and manually whitelist them to a specific container tab

I wouldn't recommend it for family or friends, but I'm happy with the trade off