ktownsend

Nice to see a clear example of doing this entirely locally on a MBP. It ran >2x faster on my M2 MBP compared to the numbers they showed for an M1. Only 23/25 of the test cases passed for me on the fine-tuned model following the README 1:1, but the speedup from fine-tuned versus off-the shelf was clear. Thanks for sharing.

Any example(s) you can point to?

Thanks for posting.

I've always had an appreciation for language, particularly concise, poignant, highly-visual word images that seem to snap you out of your mental lull in a handful of syllables.

For the most part poetry, however, was completely inaccessible to me. Uninteresting, even. It always felt too angsty, overly pretentious, seemingly trying too hard for it's own good and for my own pragmatic self to appreciate it.

Anyone can fall in love with a few easy-to-digest classics like the ending of 'The Hollow Men' (T.S. Eliot):

  This is the way the world ends
  This is the way the world ends
  This is the way the world ends
  Not with a bang but a whimper.

But nothing beyond those few textbook classics every clicked with me.

I don't know what changed a year or two back, now in my early 40s, but I decided to give poetry another chance recently, and I regret that I wasn't able to appreciate it's condensed, sometimes brutal, sometimes tender, ofttimes hit-and-miss relevance. Maybe just a lack of maturity on my part, but it's brought a lot of value to my life finding a handful of authors that resonate with me.

Most of them well known: W.H. Auden (“All we are not stares back at what we are.”), Seamus Heaney, T.S. Eliot, Baudelaire ("I am the wound and I the knife! \ I am the blow I give, and feel!")

Some unknown to me until recently, and whom I can't remember how I came across, but who I've grown to appreciate a lot: Ocean Vuong is definitely heavy on the angsty side, but God he can turn a phrase, pushing out some poignant parallel that catches you off guard. And you can forgive a bit of angst in the disorienting world we live in today.

I wish I'd have had more patience, or sensitivity, or just more of the requisite life experiences to appreciate poetry earlier, and not be ashamed of trying to be more open to it. All those emotions and ideas and feelings pared brutally down to the bone, until only a handful of sounds and syllables and some highly distilled reflection remains.

Christopher Fry wrote: "Poetry is the language in which man explores his own amazement". That expresses poetry well, for me ... at least for the authors that seem to resonate with me and my own life experiences.

If you're like me and have an interest in learning to appreciate poetry, but have no real bearings on where to start, I'd recommend 'How to Read Poetry Like a Professor' by Thomas C. Foster. He has a wonderful, witty writing style, but was an enormous help to me learning how to read and appreciate poetry.

Ordered this set of translations from Wong May just now. I'm thoroughly grounded in the Western tradition, but definitely open to a nice mental image from anywhere, and the (rather Wordsworthian) line in the subject here has a nice feel to it.

As someone with dual citizenship through marriage, and having lived 15 years in France (and speaking French at home despite English being my mother tongue), I'm amused and dismayed to discover than I'm the proverbial Karen of my adopted country ... and never had a clue before this article.

I can empathize with the gut reaction, and the title is clickbait-y, but the article does elaborate that it's not the same tired trope of privileged kid drops out of Stanford/MIT/Harvard to become startup billionaire:

> he dropped out of high school to become a poet. It would take a chance encounter during his university years — and many moments of feeling lost — for him to find that mathematics held what he’d been looking for all along.

Having taken a similar path myself -- I dropped out of high school due to a combination of factors, but was still able to get into university 6 months later -- I can empathize with mandatory education NOT being something people fit into, and where making the right career choice isn't always clear or easy.

Not everyone who 'dropped out' did it because of unlimited mom-and-dad-money and boundless professional and social opportunities, nor wears it like a badge of honor.

I get tired of the trope and survivorship bias as well, but in this case it seems like the author did still put in the hard work of learning his niche the traditional way, if in a round-about manner: through a lot of hours of study, personal investment, natural curiosity, and using the abilities he has to the best of his abilities.

Indeed, I'd hate to be the recruiter for that. I can't imagine the kind of psychological profile you'd need for something like that. The end of the article describes the psychological cost the investigation had on the investigators who were parents themselves.

But even at an algorithmic level, it seems like there are all kinds of red flags you could pull out of public posts and do a much better job of redirecting potentially victimized people to some organisations doing meaningful work to help victims.

How many categories does my online profile fit in based on all my interests, and how many hundreds of bright people are wasting their lives to discern that I happen to like A + B + C and I'm in financial bracket F, with political leaning G ... to try to show me an advert I'm probably just blocking anyway.

Thanks for taking the time to post this, and I'm sorry you had such a sh*tty start to life. I think it's important to repeat this over and over that the lines are so blurry with content online, and it's not really getting better.

The whole idea of ethical porn is something that needs more discussion, and it's an area where if you enjoy porn you should put your money where your ideals are and encourage something involving consenting adults respecting certain norms, and treating people involved with the human dignity they deserve. Paying to encourage those kinds of ethical productions is the only sensible solution, IMO. Endlessly frequenting copy-and-paste free content sites is a race to the bottom that isn't doing anyone any good.

In any case, I hope you have some people in your life who have shown you the respect you deserve as well, and been there to help when you needed it, and not just take what they could get from you whatever the cost.

Edited for clarity and typos.

Sorry if this is too long form for HN, but if you read the article you clearly have some patience as well.

I'm surprised all the comments here seem to be solely around the technical details with blockchain and crypto. It's HN, I get the technical bent and find it interesting as well, but the human side of this one is pretty rough, and it made me appreciate that there are people -- likely sorely underpaid relative to what they'd make elsewhere -- following these (sordid) threads, often at significant emotional cost to themselves.

I did appreciate the odd bits of relief like 'Bitcoin Jesus' and 'Octopus Guy':

> At one point Faruqui remembers a German official asking him, as they stood in the cold outside the Seoul hotel where they were staying, how the Americans had gotten the Koreans on board so quickly. “Oh, Octopus Guy,” Faruqui had explained. “You don’t have Octopus Guy. We have Octopus Guy.”

But mostly, humor aside, it made me wonder for the 1000th time if I'm making the right career and life choices myself in a comfy, well paid job where I'm probably near the top of the pyramid in terms of professional respect, working on problems that I think have reach and import in my narrow speciality ... but am I really solving the problems that matter? I'm not in advertising (thank whatever god you imagine), more in security lately, so the work isn't meaningless ... but I work with some brilliant people whose technical capacities I admire, and I wonder what would happen if a bit more of that gray matter was directed at solving some of the terrible problems described here?

So much money is invested in understanding the psychology of how to force better engagement and squeeze out every last penny of hapless consumers in whatever social network. What would happen if a fraction of that went into trying to focus on influencing the people making these awful, life-destroying choices and somehow (re)sensitizing them to the costs of their actions and navigating them away from that preventatively, even if the success rate is only 1-2 percent? Or identifying victims of abuse through posting patterns to try to make sure they're potentially being flagged to receive content and help they may not be able to believe even exists? How much is invested in psychological profiling to maximize profit for the most banal advertising ends, when maybe for once some of that gray matter making those algorithms could do something positive identifying patterns indicative of abuse, beyond just the current simple fig leaf approach to pretend the owners of your social network of choice cares about your well being.

I'd love to do better, and I'd take a decent pay cut if I felt I could do something for that, and maybe even get to feel a bit better as a person in this weird world as a side effect. Seems like we not only should but could do a lot more here, before this gets to the criminal investigation level.

There clearly aren't a lot of good bedfellows to be had in the region. US engagement in ME politics is rife with contradictions and blind self-interest, as are political calculations in the EU, China, or anywhere else with enough reach to matter. When you have to weigh blood, oil, military projection, and short-term regional influence in a unpredictable region, it's hard to see anyone as being on the right side of anything. At least the Russians aren't even pretending otherwise, which is sadly refreshing!

Anything involved Israel just has a depressing tendency to quickly degrade into double standards, name calling and selective vision from otherwise intelligent-seeming people I would normally respect. That's what really depresses me: the inability to have an intelligent rational discussion on anything involving Israel.

Be mercilessly critical if it's justified! I certainly am, and don't think oil is worth so much blood. But it's as necessary with Saudi Arabia, Turkey, Iran, and Palestine as it is with Israel. Just be lucid, honest and fair about it, and willing to admit you don't have all the facts or answers. Why that's so damned hard baffles me, especially from people like engineers who trade in complexity and nuance and edge cases days in and day out.

Historically, this hasn't always been the case, and the US were actively opposed to Israel's covert Nuclear program, for example, which was built with French support against the wishes of the US government at the time. The world has change significantly since then, including the French and American governments alignment of interests and weighing to pros and cons. I'm sure other people here can provide more nuance than I'm able to on the 'why', although there is a lot of emotional nonsense and baggage that comes with this particular topic, so you need to take it with a healthy dose of skepticism that there are NO blanket right and wrong people here and there is blood on everyone's hands.

Essentially, though, ignoring conspiracy theories and set in stone agendas, Israel simple represented a palatable partner in an incredibly complex, hostile part of the world and there is simple more in common between the two countries, despite their own at times competing interests. But US interests align far more with Israel, or politically moderate countries like Jordan than some others in the region.

I'm sure there is an element of playing to your electoral base TODAY, but honestly, I think it's just pragmatism at the end of the day in a region where the math can be brutal and bloody, and there are no easy choices of solutions, and real partnerships are fleeting in the region since the powerbrokers in one country or region today maybe be out of the picture tomorrow. Despite internal politics in Israel looking like a bloodsport from outside, they are at least a consistent voice in a highly turbulent region.

That's just my relatively uninformed, I'm sure biased opinion, but I don't have a fixed opinion on who the good and bad people here are, and thing there's a lot of blame in every direction, and no real will to change that on any side.

It tends to be a very small world where you get to know everyone after a certain point, or you know someone who knows someone, and jobs tend to be had by word of mouth. If you have a good reputation (which can be gained working on problems on visible open source projects), work tends to present itself. Even more so than in other fields because it is such a small world, despite the enormous amounts of money involved in semiconductors.

There aren't a lot of advertised jobs, but I think this is often because they tend to get filled via word-of-mouth and employability is rarely a problem for a decent embedded engineer.

The number of PRs and committers is what boggles my mind ... it's an endless stream, which is highly unusual for an embedded project. This is where the momentum and technical investment is happening today for open source embedded, IMHO.