5 ms·

Tech people seem to fall into the ‘perfect is the enemy of good’ trap super easily. They can’t see progress in front of their faces. Be it the internet, crypt

Tech people seem to fall into the ‘perfect is the enemy of good’ trap super easily. They can’t see progress in front of their faces. Be it the internet, crypto, EVs, reusable rockets - they’re all seen as jokes until they aren’t. AI is a juggernaut. If you can’t see it I’m kind of envious because the reality around the corner is potentially terrifying.

PaulHoule 4y ago
I think of the story from Lem’s Cyberiad about the 8 story high ‘thinking machine’ that insisted 2+2=5, got angry like the Bing chatbot when it was told it was wrong, then broke loose from its foundations and chased Trurl and Klapucius into the mountains. People will learn the hard way there is no market for machines that get the wrong answers. There are plenty of places where people will accept one kind of imperfection or another and that’s fine, but when it comes to an accounting bot that screws up your taxes it is not fine. (Funny I have seen a few chatbots that claim they are busy as soon as you tell them they are wrong about something and I wonder if that is because they’ve been trained on many examples where things really went south when somebody called out somebody’s mistake.)
- pixl97 4y ago
  >People will learn the hard way there is no market for machines that get the wrong answers. Eh, in machines there are markets for machines that get it occasionally wrong as long as the defect rate is lower and or cheaper than the human equivalent defect rate cost. So I'd say that's a really bad take on the last few hundred years of industrialization. In accounting there are really two different sets of things going on at once. There are 'the numbers' of which we'd run in a calculator, but then there is interpretation of the written rules in relation to the numbers you input. If your business is in any way complex, take your numbers to 3 different firms and see if even two come back with anything close to the same number. Hell, in the same damned firm we commonly see that two auditors will come up with different numbers and a supervisor had to look at the rule in question and make a judgement call on what they think the IRS agent would accept. Now, don't confuse that with me thinking that using ChatGPT is a good idea to do the above. We are not there yet.
- bottlepalm 4y ago
  Like I said, your hang up on perfection is blinding you to the rate of progress and potential consequences.
- morelisp 4y ago
  I share your reflection on the current state and foreseeable future effectiveness of AI, but > People will learn the hard way there is no market for machines that get the wrong answers. I fear this is wildly untrue. To take your example: The wealthy won't use the accounting bot. Everyone else won't have the time/energy/means to recoup whatever it cost them. In software development one only needs look around to see a world of "markets" - that is, profitable opportunities - for wrong answers (bad designs, useless products, orders of magnitude of inefficiency).
  - PaulHoule 4y ago
    Scary isn't it?

PaulHoule 4y ago

I think of the story from Lem’s Cyberiad about the 8 story high ‘thinking machine’ that insisted 2+2=5, got angry like the Bing chatbot when it was told it was wrong, then broke loose from its foundations and chased Trurl and Klapucius into the mountains. People will learn the hard way there is no market for machines that get the wrong answers. There are plenty of places where people will accept one kind of imperfection or another and that’s fine, but when it comes to an accounting bot that screws up your taxes it is not fine. (Funny I have seen a few chatbots that claim they are busy as soon as you tell them they are wrong about something and I wonder if that is because they’ve been trained on many examples where things really went south when somebody called out somebody’s mistake.)

pixl97 4y ago
>People will learn the hard way there is no market for machines that get the wrong answers. Eh, in machines there are markets for machines that get it occasionally wrong as long as the defect rate is lower and or cheaper than the human equivalent defect rate cost. So I'd say that's a really bad take on the last few hundred years of industrialization. In accounting there are really two different sets of things going on at once. There are 'the numbers' of which we'd run in a calculator, but then there is interpretation of the written rules in relation to the numbers you input. If your business is in any way complex, take your numbers to 3 different firms and see if even two come back with anything close to the same number. Hell, in the same damned firm we commonly see that two auditors will come up with different numbers and a supervisor had to look at the rule in question and make a judgement call on what they think the IRS agent would accept. Now, don't confuse that with me thinking that using ChatGPT is a good idea to do the above. We are not there yet.
bottlepalm 4y ago
Like I said, your hang up on perfection is blinding you to the rate of progress and potential consequences.
morelisp 4y ago
I share your reflection on the current state and foreseeable future effectiveness of AI, but > People will learn the hard way there is no market for machines that get the wrong answers. I fear this is wildly untrue. To take your example: The wealthy won't use the accounting bot. Everyone else won't have the time/energy/means to recoup whatever it cost them. In software development one only needs look around to see a world of "markets" - that is, profitable opportunities - for wrong answers (bad designs, useless products, orders of magnitude of inefficiency).
- PaulHoule 4y ago
  Scary isn't it?