5 ms·

There should be more cash prizes for solving historical mysteries

101 points by arbesman 3y ago

dang 3y ago
Recent and related: First word discovered in unopened Herculaneum scroll by CS student - https://news.ycombinator.com/item?id=37857417 https://news.ycombinator.com/item?id=37857417 - Oct 2023 (207 comments)
yieldcrv 3y ago
bounties are a good start but distributed unpaid work where even the correct solution might be rejected by the arbiter is a poor incentive method this is masked by the fact that there is participation, but it’s excluding people that value their time. people possibly more skilled in solving the problem.
- karaterobot 3y ago
  As noted in the article, challenges like these are not unusual at all in other fields, and the "distributed unpaid work" is not considered unethical, perhaps because (as with FOSS development) it is entirely voluntary. And opening it up to non-experts is the point, it's not a bug—if the non-experts can do what the experts can't, the experts are generally happy about it because they benefit from the results even if they don't win the prize.
  - yieldcrv 3y ago
    the commonality refutes nothing about my point let me merge it with your words: voluntary gambles exclude people that value their time, and waste the time of every party except the person chosen to be the winner
    - HPsquared 3y ago
      Such is life.
      - yieldcrv 3y ago
        There is a lot of active work on incentive models right now
    - karaterobot 3y ago
      So your point was just to make two value-free observations about challenges? I interpreted your comment as implying they were exploitative and unfair. Thanks for clarifying that you aren't saying that. I would make a small correction to your last message: challenges are not gambling in any sense of the word.
      - yieldcrv 3y ago
        > I would make a small correction to your last message: challenges are not gambling in any sense of the word okay, reworded since you are allergic to that one word: negative expected value games of chance exclude people that value their time who may be a lot more skilled given the validation of their time. this is a large population and there exists other incentive model for even better results. hopefully people sit down and consider competing incentive models that work for this specific market.
pvaldes 3y ago
If this knowledge is valuable, the best way to show a support for it is hiring somebody to work full time on this problem. What we have is another case of "youngsters should work in science as volunteer with their parents money, so we don't need to pay for it". In the best case one of them will produce a small result, hit the end-of-the-road signal and (most probably) will move away. The other try and quit after investing X hours by the small chance of future reward and none of the social benefits that came with real work. My bet is that none of them will use that cash to pursue a career to turn into a professional in history. The real experts are getting very old and would need replacement, but if the job offer is the equivalent to "buy lottery tickets with your money and time with the hope to win 10k", it just can't compete. The real lottery requires only money and offers a bigger reward without effort. Is a non efficient way to generate knowledge, because you are burning most of the talented people in the process, and is also anti-economical as a society.
- benbreen 3y ago
  "If this knowledge is valuable, the best way to show a support for it is hiring somebody to work full time on this problem." Why not have both? I.e. academics and students dedicated to the fairly monastic career of being a professional historian (which I am), but also prizes which seek to bring in experts from other fields with different skillsets who may approach the problem in an original way. Obviously there is a limited amount of funding for anything. But surely some of the billions being donated every year to the general funds of elite universities like Harvard would be better spent setting up specific, goal-oriented prizes modeled on the Vesuvius Challenge.
- bruce511 3y ago
  >> If this knowledge is valuable, the best way to show a support for it is hiring somebody to work full time on this problem. Best for whom? I'm not sure it's best for the employer since they then live or die based on the abilities, insight and inspiration of that one employee. Equally, I'm not sure it's best for the employee who becomes singularly focused on that one problem to the exclusion of all others. If they want to diverge down an alternate path for a bit they (likely) cannot. Of course in one sense work-for-prize is not efficient. We may as well argue that FOSS programing is not efficient. But FOSS exists not because of effeciency, but because it allows people to scratch an itch (usually unpaid.) Prizes allow specific goals to be prioritised within that community. Say, for example, an expert at ocean currents notices a prize for decoding a historical text. Sounds like a fun hobby. Figures out that it contains tide tables. Employment is clearly a good thing. Scanning in old archives could easily be done by an employee. But there's a place in the world for more than just formal employment. The existence of prizes does not negate employment. Rather it might attract folk into a "hobby" of sorts.
paulpauper 3y ago
But I think this news is important for another reason, too. It’s strong evidence that cash prizes — even relatively modest ones like $10,000 — are effective in producing new historical knowledge. $10k modest? I think that is more than most bug bounties, even quite serious ones. Part of the reason the payouts are low or why this is not done more often, is because there is too much work involved relative to to the opportunity cost. It's basic econ. The $40k prize for deciphering a word was an especially 'good deal' but most prizes are worse. It's also cheaper to have an expert on payroll doing work everyday than one-time payouts.
- queuebert 3y ago
  I would like to live in a world where solving an ancient mystery pays better than what a realtor gets for selling a single house.
  - choxi 3y ago
    How much is that? The grand prize for the Vesuvius Challenge is $700k
    - TylerE 3y ago
      It’s fairly significant. 6% is standard, and million+ dollar homes are not exactly rare these days.
  - midasuni 3y ago
    Average seller in the uk gets paid £1k for online only, £3k if you want somewhere with a physical presence. That’s far too much IMO.
- TylerE 3y ago
  Well, I mean, that’s probably about 2 weeks of work at contractor rates. Not that much in the grand scheme of things. Especially when you’re only paying for wins, not attempts
captainpicktart 3y ago
[flagged]
- rthomas6 3y ago
  Let me know when the Culture Minds are ready
dekhn 3y ago
CROATOAN
bjackman 3y ago
I don't think the examples given by the author are good ones (although I don't entirely reject the conclusion). Doing what is essentially advanced computer vision is a very different type of problem from decoding Linear A. The Herculaneum Scrolls are a case where we have a relatively self-contained source of information and we "just" need technical skills, which historians probably lack, to extract it. Deciphering Linear A or the Indus Valley script is a case where the information is simply not available in any given artifact (or any set of artifacts that someone can study in a short period). No amount of technical prowess or brilliance from an "outsider" is gonna magically add that information. The only way you can hope to unravel that mystery is with deep expertise in the archaeologal and linguistic CONTEXT. That takes years or decades to build up. Which irequires exactly the opposite kind of funding: stable employment in an opportunity-rich environment. Not a one-off prize.
vintermann 3y ago
Cash prizes are not a good way of funding most things. If you can't afford to spend time on something that will most likely fail, a cash prize won't change that. Also, if the cash prize is what motivates people, they'll be less willing to cooperate and share progress.
- flir 3y ago
  To be honest I think the prestiege and the CV line item is the main motivation for the kids that solve these problems (beyond it just being an interesting problem of course). The existence of the cash prize probably works more as cheap PR, making sure news of the competition reaches them. (I'm saying "$10k to your favourite charity" would work almost as well as "$10k for you").
NoMoreNicksLeft 3y ago
Is that supposed to be a C, or a terminal Sigma? I don't read Greek well at all, let alone any variety of ancient Greek, but that's not a letter in the alphabet the way I learned it.

101 points by arbesman 3y ago

dang 3y ago

Recent and related: First word discovered in unopened Herculaneum scroll by CS student - https://news.ycombinator.com/item?id=37857417 https://news.ycombinator.com/item?id=37857417 - Oct 2023 (207 comments)

yieldcrv 3y ago

bounties are a good start but distributed unpaid work where even the correct solution might be rejected by the arbiter is a poor incentive method this is masked by the fact that there is participation, but it’s excluding people that value their time. people possibly more skilled in solving the problem.

karaterobot 3y ago
As noted in the article, challenges like these are not unusual at all in other fields, and the "distributed unpaid work" is not considered unethical, perhaps because (as with FOSS development) it is entirely voluntary. And opening it up to non-experts is the point, it's not a bug—if the non-experts can do what the experts can't, the experts are generally happy about it because they benefit from the results even if they don't win the prize.
- yieldcrv 3y ago
  the commonality refutes nothing about my point let me merge it with your words: voluntary gambles exclude people that value their time, and waste the time of every party except the person chosen to be the winner
  - HPsquared 3y ago
    Such is life.
    - yieldcrv 3y ago
      There is a lot of active work on incentive models right now
  - karaterobot 3y ago
    So your point was just to make two value-free observations about challenges? I interpreted your comment as implying they were exploitative and unfair. Thanks for clarifying that you aren't saying that. I would make a small correction to your last message: challenges are not gambling in any sense of the word.
    - yieldcrv 3y ago
      > I would make a small correction to your last message: challenges are not gambling in any sense of the word okay, reworded since you are allergic to that one word: negative expected value games of chance exclude people that value their time who may be a lot more skilled given the validation of their time. this is a large population and there exists other incentive model for even better results. hopefully people sit down and consider competing incentive models that work for this specific market.

pvaldes 3y ago

If this knowledge is valuable, the best way to show a support for it is hiring somebody to work full time on this problem. What we have is another case of "youngsters should work in science as volunteer with their parents money, so we don't need to pay for it". In the best case one of them will produce a small result, hit the end-of-the-road signal and (most probably) will move away. The other try and quit after investing X hours by the small chance of future reward and none of the social benefits that came with real work. My bet is that none of them will use that cash to pursue a career to turn into a professional in history. The real experts are getting very old and would need replacement, but if the job offer is the equivalent to "buy lottery tickets with your money and time with the hope to win 10k", it just can't compete. The real lottery requires only money and offers a bigger reward without effort. Is a non efficient way to generate knowledge, because you are burning most of the talented people in the process, and is also anti-economical as a society.

benbreen 3y ago
"If this knowledge is valuable, the best way to show a support for it is hiring somebody to work full time on this problem." Why not have both? I.e. academics and students dedicated to the fairly monastic career of being a professional historian (which I am), but also prizes which seek to bring in experts from other fields with different skillsets who may approach the problem in an original way. Obviously there is a limited amount of funding for anything. But surely some of the billions being donated every year to the general funds of elite universities like Harvard would be better spent setting up specific, goal-oriented prizes modeled on the Vesuvius Challenge.
bruce511 3y ago
>> If this knowledge is valuable, the best way to show a support for it is hiring somebody to work full time on this problem. Best for whom? I'm not sure it's best for the employer since they then live or die based on the abilities, insight and inspiration of that one employee. Equally, I'm not sure it's best for the employee who becomes singularly focused on that one problem to the exclusion of all others. If they want to diverge down an alternate path for a bit they (likely) cannot. Of course in one sense work-for-prize is not efficient. We may as well argue that FOSS programing is not efficient. But FOSS exists not because of effeciency, but because it allows people to scratch an itch (usually unpaid.) Prizes allow specific goals to be prioritised within that community. Say, for example, an expert at ocean currents notices a prize for decoding a historical text. Sounds like a fun hobby. Figures out that it contains tide tables. Employment is clearly a good thing. Scanning in old archives could easily be done by an employee. But there's a place in the world for more than just formal employment. The existence of prizes does not negate employment. Rather it might attract folk into a "hobby" of sorts.

paulpauper 3y ago

But I think this news is important for another reason, too. It’s strong evidence that cash prizes — even relatively modest ones like $10,000 — are effective in producing new historical knowledge. $10k modest? I think that is more than most bug bounties, even quite serious ones. Part of the reason the payouts are low or why this is not done more often, is because there is too much work involved relative to to the opportunity cost. It's basic econ. The $40k prize for deciphering a word was an especially 'good deal' but most prizes are worse. It's also cheaper to have an expert on payroll doing work everyday than one-time payouts.

queuebert 3y ago
I would like to live in a world where solving an ancient mystery pays better than what a realtor gets for selling a single house.
- choxi 3y ago
  How much is that? The grand prize for the Vesuvius Challenge is $700k
  - TylerE 3y ago
    It’s fairly significant. 6% is standard, and million+ dollar homes are not exactly rare these days.
- midasuni 3y ago
  Average seller in the uk gets paid £1k for online only, £3k if you want somewhere with a physical presence. That’s far too much IMO.
TylerE 3y ago
Well, I mean, that’s probably about 2 weeks of work at contractor rates. Not that much in the grand scheme of things. Especially when you’re only paying for wins, not attempts

captainpicktart 3y ago

[flagged]

rthomas6 3y ago
Let me know when the Culture Minds are ready

dekhn 3y ago

CROATOAN

bjackman 3y ago

I don't think the examples given by the author are good ones (although I don't entirely reject the conclusion). Doing what is essentially advanced computer vision is a very different type of problem from decoding Linear A. The Herculaneum Scrolls are a case where we have a relatively self-contained source of information and we "just" need technical skills, which historians probably lack, to extract it. Deciphering Linear A or the Indus Valley script is a case where the information is simply not available in any given artifact (or any set of artifacts that someone can study in a short period). No amount of technical prowess or brilliance from an "outsider" is gonna magically add that information. The only way you can hope to unravel that mystery is with deep expertise in the archaeologal and linguistic CONTEXT. That takes years or decades to build up. Which irequires exactly the opposite kind of funding: stable employment in an opportunity-rich environment. Not a one-off prize.

vintermann 3y ago

Cash prizes are not a good way of funding most things. If you can't afford to spend time on something that will most likely fail, a cash prize won't change that. Also, if the cash prize is what motivates people, they'll be less willing to cooperate and share progress.

flir 3y ago
To be honest I think the prestiege and the CV line item is the main motivation for the kids that solve these problems (beyond it just being an interesting problem of course). The existence of the cash prize probably works more as cheap PR, making sure news of the competition reaches them. (I'm saying "$10k to your favourite charity" would work almost as well as "$10k for you").

NoMoreNicksLeft 3y ago

Is that supposed to be a C, or a terminal Sigma? I don't read Greek well at all, let alone any variety of ancient Greek, but that's not a letter in the alphabet the way I learned it.