• 0 Posts
  • 380 Comments
Joined 6 months ago
cake
Cake day: December 6th, 2024

help-circle
  • Well, that’s the thing: LLMs don’t reason - they’re basically probability engines for words - so they can’t even do the most basic logical checks (such as “you don’t advise an addict to take drugs”) much less the far more complex and subtle “interpreting of a patient’s desires, and motivations so as to guide them through a minefield in their own minds and emotions”.

    So the problem is twofold and more generic than just in therapy/advice:

    • LLMs have a distribution of mistakes which is uniform in the space of consequences - in other words, they’re just as likely to make big mistakes that might cause massive damage as small mistakes that will at most cause little damage - whilst people actually pay attention not to make certain mistakes because the consequences are so big, and if they do such mistakes without thinking they’ll usually spot it and try to correct them. This means that even an LLM with a lower overall rate of mistakes than a person will still cause far more damage because the LLM puts out massive mistakes with as much probability as tiny mistakes whilst the person will spot the obviously illogical/dangerous mistakes and not make them or correct them, hence the kind of mistakes people make are mainly the lower consequence small mistakes.
    • Probabilistic text generation generally produces text which expresses straightforward logic encodings which are present in the text it was trained with so the LLM probability engine just following the universe of probabilities of what words will come next given the previous words will tend to follow the often travelled paths in the training dataset and those tend to be logical because the people who wrote those texts are mostly logical. However for higher level analysis and interpretation - I call then 2nd and 3rd level considerations, say “that a certain thing was set up in a certain way which made the observed consequences more likely” - LLMs fail miserably because unless that specific logical path has been followed again and again in the training texts, it will simply not be there in the probability space for the LLM to follow. Or in more concrete terms, if you’re an intelligent, senior professional in a complex field, the LLM can’t do the level of analysis you can because multi-level complex logical constructs have far more variants and hence the specific one you’re dealing with is far less likely to appear in the training data often enough to affect the final probabilities the LLM encodes.

    So in this specific case, LLMs might just put out extreme things with giant consequences that a reasoning being would not (the “bullet in the chamber” of Russian roulette), plus they can’t really do the subtle multi-layered elements of analysis (so the stuff beyond “if A then B” and into the “why A”, “what makes a person choose A and can they find a way to avoid B by not chosing A”, “what’s the point of B” and so on), though granted, most people also seem to have trouble doing this last part naturally beyond maybe the first level of depth.

    PS: I find it hard to explain multi-level logic. I supposed we could think of it as “looking at the possible causes, of the causes, of the causes of a certain outcome” and then trying to figure out what can be changed at a higher level to make the last level - “the causes of a certain outcome” - not even be possible to happen. Individual situations of such multi-level logic can get so complex and unique that they’ll never appear in an LLMs training dataset because that specific combination is so rare, even though they might be pretty logic and easy to determine for a reasoning entity, say “I need to speak to my brother because yesterday I went out in the rain and got drenched as I don’t have an umbrella and I know my brother has a couple of extra ones so maybe he can give one of them to me”.


  • Sounds to me like you’re doing the fun part of the job - “solving challenging problems” - without having to do the vast majority of the work (which is seldom as much fun), such as making it suitable for actual end users, integration with existing systems and/or migration, maintaining it during its entire life-cycle, supporting it (which for devs generally means 3rd level support) and so on.

    So not exactly a typical environment from which to derive general conclusions about what are the best characteristics for a professional in software engineering in general.

    Mind you, I don’t disagree that if what you’re doing is basically skunworks, you want enthusiastic people who aren’t frozen into a certain set of habits and technologies: try shit out to see if it works kind of people rather than the kind that asks themselves “how do I make this maintainable and safe to extend for the innevitable extra requirements in the future”.

    Having been on both sides of the fence, in my experience the software that comes from such skunkworks teams tends to be horribly designed, not suitable for production and often requires a total rewrite and similarly looking back at when I had that spirit, the software I made was shit for anything beyond the immediacy of “solving the problem at hand”.

    (Personally when I had to hire mid-level and above devs, one of my criteria was if they had already been through the full life cycle for a project of theirs - having to maintain and support your own work really is the only way to undrestand and even burn into one’s brain the point and importance of otherwise “unexplained” good practices in software development and design).

    Mind you, I can get your problem with people who indeed are just jobsworths - I’ve had to deal with my share of people who should’ve chosen a different professional occupation - but you might often confuse the demands and concerns of people from the production side as “covering their asses bullshit” when they’re in fact just the product of them working on short, mid and long term perspectives in terms of the software life-cycle and in a broader context hence caring about things like extensability, maintenability and systems integration, whilst your team’s concerns end up pretty much at the point were you’re delivering stuff that “works, now, in laboratory conditions”. Certainly, I’ve seen this dynamic of misunderstandings between “exploratory” and “production” teams, especially the skunkworks team because they tend to be younger people who never did anything else, whilst the production team (if they’re any good) is much more likely to have at least a few people who, when they were junior, did the same kind of work as the skunkworks guys.

    Then again, sometimes it really is “jobsworths who should never have gone into software development” covering their asses and minimizing their own hassle.






  • Ah, no concreted metrics for efficiency and delivery of results.

    Explains why you prioritize employees who have fun on the job rather than efficient professionals who are there to do a job well done - you can’t really like to like compare with other teams (much less the broader industry) when it comes to delivering objectives because it’s all open ended and unique, so you really don’t know for sure which kind of employee is more effective but you do know for sure which kind is more fun to work with, hence you prioritize what you can measure - a fun team - not what is more effective and efficient.

    Most work out there in software development is not “cracking interesting problems for fun without a strict timeline”, it’s “integrate new functionality into an existing massive custom-made system, which has at least 3 different styles of programming and software design because different people have worked on it over the last 8 years and only not a complete mess of spaghetti code if you’re lucky” - not really the kind of work were Enthusiasm lasts long, but it still has to be done and sometimes, millions, tens of millions and even hundreds of millions in yearly revenue of some company or other rides in doing that job well and in a timelly fashion.

    Don’t take this badly, but from where I’m standing you’re in the playground sandbox of software engineering. No doubt it’s fun and even an environment others would love to be able to work in, it’s just not the place for professionals and doesn’t really reflect most of the software development being done out there, so not exactly a representative environment for determining what kind of professionals are suitable for the wider industry.


  • Look mate, I’ve been in Software Development for almost 3 decades, mainly in the Technical careed path (did some Project Management but, frankly, it’s not my thing) and all the way to Technical Architect, in 3 different countries and most of it as a contractor, so I worked in quite a number of companies and work environment.

    (I’m not trying to pull rank here, just showing that I’ve seen a lot)

    In my experience, things like Enthusiasm are what bright eyed naive junior developers have: they’re like me as a teen in the swiming pool having learnt to swim by myself and never having had lessons - intense strokes trowing water all over the place but moving very little for all that effort, or in other words lots of effort with little in the way of results.

    Worse, Enthusiasm doesn’t last forever and, further, most of the work than needs to be done is not exactly stimulating (if it was fun, people wouldn’t have to pay money to others for doing it).

    People who get at least some enjoyment of their work are good to have (and I’m lucky that after all these years I still get those moments of great enjoyment when at the end of doing something insanelly complex it all works), but in the real world most work that needs to be done is needed but boring so fun in that kind of task by itself won’t be enough, plus such people are actually uncommon beyond the bright eyed young things, so if you want somebody who will actually deliver you results (rather than work a lot to achieve little) and you’re not a prestigious company (say, like Google, which leverages their brand recognition to pull in such bright young things by the bucket load and drip them out drained of on the other side) and can’t pay well above average, you’re highly unlikely to get those kinds of people.

    What you really want is people who have things like professional pride: they want to do a good job because they see themselves as professionals and feel a professional responsability to deliver good results in an efficient way that doesn’t hinder the work of others.

    I’ve seen over the years people with your perspective heading Startups or teams within small companies, and invariably they end up with unproductive teams filled with inexperienced people making all the mistakes in the book (and inventing new ones), enthusiastically. Maybe the people seeking such workers should’ve asked themselves what their real objective is in that: is it deliver the results needed by the company so that it prospers and grows or is it the pleasure of being surrounded by people having fun.


  • IMHO, in Software Development it’s a good idea for a candidate to ask about the project, if only because any good professional would want to know if they’re a good fit or not.

    Mind you, that makes sense in the Technical interview rather than with HR - no point in asking about what are the practical professional details of the work you will be doing from a person who doesn’t really have a clue (the HR person) when you know you will be facing an actual professional peer in a technical interview who knows the work that needs to be done in your terms and with the level of detail and understanding only domain professionals have.

    In my experience doing the Technical Interview side of things (and most of my career I was a Contractor - so a Freelancer - which is hardly a “company man” with a rosy view of my relationship to them or somebody who thinks people work for fun), people who don’t ask about the project during the Technical Interview tend to as the interview proceeds end up get revealed as technically weak: an experienced “Engineer” would want to make sure they’re well matched to the kind of work they’re be doing (as well as, in my experience from the other side of the interviewing table, spot the messy fucked up situations before you take the contract so that if you can avoid ending in such disfunctional environments).


  • I mean, the whole “this is your second family” or “you should be proud of were you work” thing isn’t bad if they’re similarly dedicated to their employees welfare, for example “no questions asked sick days off” or maybe even more relevant in Tech, sizing the team to the work that need to be done in a project rather than expecting constant unpaid overwork from employees (rather than just once in a while).

    The problem, as emphasized by the OP, is that they expect employees to invest themselves in the company without the company investing in employees.

    There apparently are some companies out there which are almost like a second family, you know, the kind of place were they hear that your grandmother died and give you a week paid leave no questions asked to “deal with your loss”, but most aren’t at all like that - they treat employees as disposable cogs whilst expecting that the employees respond back by being dedicated to the company.


  • It’s either a business relation on both sides or it’s a personal relation on both sides.

    I was in Tech in Europe through the transition from when employees were people and the company was loyal to them and expected loyalty to the company in return (the age of lifetime employment), to the world we live in now were employees are “human resources”, and for a great part of that period there was this thing were most employers expected employees to stay with the company whilst the company needed them and be dedicated to the company, whilst in return they treated employees as a business relationship with (in Tech) some manipulative “fake friendship” stuff thrown in (the ultimate examples: company paid pizza dinner when people stay working on a project till late, or the yearly company party, rather than, you know, paying people better or sizing the team to fit the work that needs to be done rather than relying on unpaid overwork) - still today we see this kind of shit very obviously and very purposefully done in places like Google.

    Of course the “humour” part here is that plenty of managerial and HR people in companies still expect that employees are loyal to the company even all the while they treat them as disposable cogs who it’s fine to exploit without consideration for their feelings or welfare - or going back to the first paragraph of this post: they relate to employees as a business relationship whilst expecting the employees related to the company as a personal relationship (often a “second family”).





  • Yeah, well, Asset Owners and the Speculators who caused the 2008 Crash got saved by Governments all around the World with public money and that was done by taking money away from those whose income comes from their Work (notice how salaries keep getting less and less able to cover living costs all the while asset prices keep beating records).

    What I learned from the 2008 Crash - being right smack in the middle of the Industry which made it happen and spending the subsequent years observing what Governments were doing to “recover” from it and trying to figure out “how did this happen” - was that politicians don’t work for most people, they work for a handful of people, and when push comes to shove they’ll sacrifice the rest to make sure those few keep on accumulating riches.

    The fishy speculation that played around the line separating Legal from Fraud has in the aftermath of the 2008 Crash fully became Fraud and more, once the industry figured out that politicians from the main parties had their backs, which is why we find ourselves were we are now. This was especially so in places like the US and UK were power is a duopoly which just alternates between two groups of politicians who all frequent the same social circles and send their children to the same private schools.

    It’s a big group and we ain’t in it.



  • They’re citizenships, not nationalities.

    Israel does this almost unique thing in the World which is split Citizenship from Nationality, with those who have the former only having the latter if they’re Jews (constitutionally only a Jew can be an Israeli national). Further Israeli Citizenship is broken into Jewish Israeli and Non-Jewish Israeli, with the latter having less rights than the former (for example, Non-Jewish Israeli Citizens require “authorization” from local authorities to go live in certain parts of the country).

    Israel was set up as an Apartheid state from the very beginning.

    That said the other poster made a mistake: Palestinians are not Israelis - not even Non-Jewish Israeli Citizens - and have no rights at all in Israel. Further, there is no right to Israeli Citizenship by being born in Israel (unless you’re a Jew: any Jew, anywhere in the World, has a right to Israeli Nationality, which also gives them Jewish Israeli Citizenship).

    This has interesting effects such as people whose families have lived in Jerusalem for generations not having Israeli Citizenship because it was denied to them even though they were born there and lived there their whole lives. Of course, all of these people are not Jewish as all Jews constitutionally have a right to Israeli Nationality.

    So By Law Israel has 3 different classes of people living there, in decreasing order of the rights they have:

    • Jews, which by law are all Israeli Nationals (a Jew can literally arrive in Tel-Aviv, ask for Israeli Nationality and get it) and hence are also Jewish Israeli Citizens.
    • Non-Jewish Israeli Citizens
    • Non-Israeli Citizen, which includes all Palestinians even those born in the territory of Israel as Israel can and frequently does simply refuse to give them even just Non-Jewish Israeli Citizenship.


  • Aceticon@lemmy.dbzer0.comtomemes@lemmy.worldBenefit of the hindsight
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    3 days ago

    First, apologies for the wall of text. I did try to make it short, but it’s a multi-step complex subject.

    What you wrote about possession and ownership makes sense in the physical world, were there can be only be one person who has possession of something - I give you the seashell I have in my possession and it’s now in your possession and not in my possession anymore.

    In the digital world, on the other hand, there can be infinite people in possession of exactly the same works because it can be copied at no cost.

    This is why we end up with the headfuck which is Copyright Legislation and terms like “ownership” and “licensing” having very specifical legal meanings: whilst possession in the physical domain is unique and stupidly simple to determine (though, as you correctly pointed out, less so “ownership”), in the non-physical domain if you want to preserve that unique possession (or at least limit the number of people having possession) you end up with the anti-natura construct which is Intelectual Property Legislation.

    Now, NFTs are a two step mechanism for asserting possession - they’re not the asset itself but rather a token which itself is possessed and whose possession gives rights over an underlying asset.

    The problem of unique possession of the NFT itself is techically solved by NFTs. However NFTs have no mechanism whatsoever to enforce any rights on the underlying asset - that part is delegated to other, non-technical mechanisms, which for underlying assets which are non-physical goods, invariably means Copyright Legislation and Contract Law as those are the only ways to control rights over infinitelly-copiable things (well, there’s DRM, but NFTs don’t do that).

    As you pointed out NFTs, as a standardized token representing possession, deliver various practical benefits, which I would say are similar to those that Stocks deliver as tokens representing part-ownership of a business.

    However:

    • For Stocks there are a lot of legally mandated business ownership rights linked to Companies Limited By Shares (or whatever is the local legal name were you live for those companies were ownership is determined by owning shares of the company) which define most of the rights that possession of Stocks give (with even tighter rules for market traded companies)
    • For NFTs there are no guaranteed legal rights on the underlying asset associated with having them: the rights on the underlying asset of an owner of an NFT start and stop with what’s in the Contract and those contracts aren’t even standardized.

    The problem is that all together the NFT architecture is not standardized because that second link of the chain - the Contract linking the NFT to the underlying non-physical asset - is non-standardized hence you don’t really know what you get if you buy a random NFT without reading the Contract for that specific NFT.

    None of this is a problem if all you want is a cheap license to use a copyrighted work and in the process help the artist by paying them directly without paying middlemen (which seems to be your use case).

    This is a problem for those who think they can just buy an NFT as a reselable investment asset, since the real value of an NFT is down to the nitty gritty details of the Contract associated with that NFT and whether one can actually enforce that Contract. It’s this part that anchors all the scams around NFTs.

    That said, the idea of using NFTs as a technological base for markets to make it simple and easy for consumers to buy works such as music directly from artists in a world were there is Copyright Legislation has merit, it’s just as an investment that NFTs are problematic.


  • I used to work in Tech back in the late 90s, during the first bubble and up to the Tech Crash. I also worked in both Investment Finance and Tech Startups much more recently.

    I can’t even begin to describe just how angry, disgusted and dissapointed this unholly intersection between Tech and Hyperspeculative-Finance of the XXI century makes me.

    The whole spirit in pretty much the entire domain of Tech back in the 90s was completely different from this neverending bottomless swamp of crap we have in the supposedly innovative parts of Tech.

    Ever since the sleazy slimeballs who saw from the first Tech bubble that there were massive opportunities to use Tech-mumbu-jumbo to extract money from suckers started (immediatly after the Crash) trying to pump the Net bubble back up (they even called it Web 2.0) that the old spirit of innovation for the sake of improving things of the old days in Tech was crushed and replaced by the most scammy, fraudulent, naked greed imaginable.

    After my time in Finance (which, curiously, also involved a Crash in the Industry I was working in) I started describing Tech Startups as “The Even Wilder Wild West of Speculative Finance”.