The NEXT BIG THING in Behavioural Economics
Вставка
- Опубліковано 11 чер 2024
- Noise is an often overlooked concept in behavioural science, and trying to solve it could have a huge payoff for society. What do you guys think of Noise?
Buy Noise here: amzn.to/3t0RckP
Great book summaries here: Shortform: www.Shortform.com/pete
My Website: petejudo.com
Follow me:
Behavioral Science Instagram: @petejudoofficial
Instagram: @petejudo
Twitter: @petejudo
LinkedIn: Peter Judodihardjo
Good tools I actually use:
Shortform: www.Shortform.com/pete
Ground News: ground.news/Pete
software engineer here who has interviewed hundreds of candidates for job positions: for the love of god, don't trust such things to algorithms: it's very easy to make algorithm consistent (for example: if candidate's age is odd then hire, else don't hire), but it doesn't make it good. algorithms only perform well, if *ALL* aspects of the problem are known, perfectly well understood and accurately measurable. that's rarely the case. it's definitely not the case for job interviews and anything involving psychology or emotions.
Yes, algorithms are less noisy, but there are many things that they can't do, and they are also subject to the bias of the creator. They are not nesecarrily objective just because they aren't noisy, they can be biased.
I read Kahneman's Thinking Fast and Slow. He actually explained his opinion there too. In a nutshell, humans have never consistently beaten algorithms since Paul Mehl's 1953 study. Not even once.
But Kahneman also advised against hiring or not someone based on a single criterium or a feeling. So a combination of both techniques could be marvelous.
Nah, uh-uh. I don't believe you. You say you're a software engineer and then proceed to make a prefectly sensible and well thought-out statement about psychology? Nah, man. You may have fooled others, but not me. I call bullshit.
IT recruiter here. Behavioral team interviewing was designed to reduce biases, but it doesn't demand that the interviewer biases are examined. Great point highlighted in this video.
A good analogy for noise would be like all things being equal an archer aiming in an identical way would hit the same spot every time, but wind acting randomly in a way that cannot be compensated for adds slight variation to where the arrow hits. But if say 50% of the variation in wind fluctuation can be compensated for, that would greatly reduce the variation in outcome thus allowing us to better see the trend in some other variable, say aiming or bow quality. The unexplained variance in the wind is noise, but reduce the unexplained variance and you reduce the noise.
You can find an interviewer example to illustrate pretty much anything in behavioral science. HR is all about arbitrary decisions with no accountability based on poor insight and pop-psychology. We all have stories.
If hr is conducting interviews, that is the problem. Recruiting is a marketing funtion, not hr.
Before the video ended I already knew this was going to go with algorithms, it has been shown we give too much credit to algorithms, people are more nuanced and that's also beneficial
Algorithms are subject to pattern noise given the datasets they are given, and the exact calculations programed to be made in the process. We have seen that Data sets can be flawed or deliberately misleading with Tessier-Lavigne, Stapel, and Gino, and that is sure to cause Pattern noise that is more akin to straight up bias and would that be anymore fair than people making pattern noise decisions? Lets think very carefully before we even consider handing over decision making processes to algorithms, people can be held accountable, an algorithm can't be, and i doubt that those that made it would be either.
Just found this channel and think its great! Well done and researched. I'd love to see longer format updates... Cheers.
I work at a risk analysis company and this book greatly impacted our work, I actually prefer this over "Think Fast, Think Slow" since it is way more focused
Thinking Fast and Slow is in fact about reducing impulsivity through deliberate thinking/decision making.
Pete, your understanding of this universe and sacrifices endured is much appreciated. Thank you.
What sacrifices and understanding of the universe does he have?
Pete, examples you mention here are in fact biases from I/O and/or Individual Psychology and systems in general, even hiring systems and tests, are built to minimize effects from these biases in decision making (horns or halo effect or influence of mood on decision making). I think the various branches of Behavioural Science/Economics, I/O, Cognitive Psychology need to integrate the terminologies so that the same concepts are not labelled differently.
Hiring decisions will always be biased. I'm a headhunter and I'm biased against people with substandard communication skills. The problem is that not every job requires candidates to possess excellent communication skills.
your narration is excellent
Are the first two even considered noise? Because the variance in both cases can be explained and estimated. Some raters being consistently harsh or lenient especially doesn't seem like it has much to do with the topic. There are also many good studies showing that on average these random blips don't really matter especially when compared to the heritable factors making some people more mentally unstable. These kinds of priming studies with judges and other people just aren't very good and don't replicate. I wouldn't trust the authors as far as I could throw them. It's also very funny seeing someone like Sunstein talk about something that could be mitigated with objective and more measures, like IQ tests, when he's a full fledged supporter of the civil rights regime which constantly works to make such measures illegal.
"IQ tests" lol you are a clown
It's probably pretty unlikely that random noise has more of an influence than heritable personality differences in personnel. But when you use words like costly, costly in what sense? Companies have many competing individuals within them and aren't very intelligent relative to their aggregate personnel. You could say if you reduce the mental instability of your hiring staff by x%, you will have y% economic improvement, but using unfair and other value loaded words just seems like a way to smuggle in a particular morality. Like you could do good science of a "bad" economic system where the supporters of said system are more prone to doing bad science due to their character, but that would say nothing about proving your moral view of the system. I guess you can say without good predictions, people can't necessarily even get what they purportedly want
"It's probably" lol
I thought thinking fast and slow got debunked? We found out that its not as simple as system 1 and system 2 thinking. Im very skeptical of Daniel
Can you show „where“ it got debunked?
@@Farmer1188 just search for contradictory theories. Obviously his model isn't perfect but it is useful for conceptualizing how the brain works. The thing is Daniel knows his theory isnt accurate but he doesn't posit it that way So at least hes being intellectually honest.
Yeah I think he has a video criticizing it
Much of Sunstein's work is nonsense too. He also does considerable work to shift semantics/the goal posts for his own selfish ends which is why he's so popular.
@@supersonic118boi8I'd say that thinking fast and slow, possibly the most influential psychology book ever, does not have a " theory " but what Daniel Dennett might call a "thinking tool", which embodies and summarize the findings about the various limits of our heuristics and the effort that we need to do to bypass said heuristics (when needed).
System 1 and system 2 are such thinking tool. I remember Kahneman's disclaimer about the two system, clarifying that they were not real in the least , that he was only using precisely our heuristics/love for narrative and characters to build these images which _functionally_ approximate our most typical cognitive behaviour.
Maybe with "debunked" you are referring to the chapter, in the original edition, referring to studies which haven't been replicated and in fact I believe they've been retracted. I understand the chapter is not there anymore , and the author have done enough mea culpa throughout the years for trusting a single group's experiments; that was a non common sign , in fact , of intellectual honesty.
The chapter was about one class of our many silly cognitive instinct, so the general tool is still valid
Really like your clear, low-_noise_ style of presentation. :-)
That said, what's going on with the choppy audio? Could it be one of those AI rewrite tools or artifacts of manual cutting? Either way there's a lot of glitching between words that I find pretty distracting, especially using headphones.
When I make cuts I zoom in to make sure I cut after the wave goes down completely. He is cutting before that which is making that noise. That is the sort of thing that makes a huge difference in the end but its boring AF to do. I assume he is doing it himself and just didn’t study audio like I did at Berklee College of Music. If he hires a pro that will be what they do. Those tedious things make the difference in the end but it really is a pain in the ass and takes hours.
My fav channel atm xx
Comment for the algorithm. This is a great video and talks about a great topic.
Small editing mistake at 1:50 or so
What could be more serious than a 'simple' game of golf?
Are there downsides to removing noise? Loss of serendipity, forging of homogeneity?
Serious question - why is this considered behavioral psychology? and How do you account for all the variables after someone is hired? Onboarding, culture fit , direct management etc... thank you, new to your channel enjoying it
8:47 - As any large enough sampling
"Jane" seems to be interested in the way that her candidate is sitting which may be contributing to her experiencing noise. See: 6:33 to 6:36 and watch where Jane looks and her eye contact and body language after dropping her gaze.
slight editing mistake around 1:50
Pete how to find out our Strengths ? Especially in the workplace that can be beneficial for us to understand & then promote ourself for becoming more Valuable..
Try a bunch of things and stop doing the things that aren't working out well.
As a researcher who uses critical theories, I'll just say that with all the occasions of computers and algorithms ending up being racially biased I don't know if I'd trust them to eliminate noise without still being biased. But it may still be better than a biased person who comes in with a lot of noise against certain candidates.
The hiring process is generally intentionally discriminatory. While seeking the best candidates, it is presumed unfit candidates will be rejected. If this changes then the is no point to having an interview process.
Additionally, over reliance on formulas, computing, and other conveniences simply for convenience and efficiently reduces the competence of the people who use them, or worse, creates systems that reward incompetence.
This said, racism sucks. It developed as a social construct thousands of years ago. But it also served a variety of purposes. Today it's used as a collective/corporate means of creating a "class consciousness" to affect a change. The results of that class creation has resulted in a variety of "biases" in reality (meaning the data is neither a noise nor artificial selection) that are experienced by "others" and expressed in behaviors, such as creation of or acceptance of stereotypes. Again, racism sucks. Using it to stoke hate to invoke a revolution sucks just as bad and shows the incompetence of the philosophy driving/justifying it.
🤣
so fckn beautiful video i cry
I read it but the language is not easy to understand 😢
❤❤❤
I don't see any real difference between Noise to Bias & Priming (or other iterations of entropy). Just a new name
Comment for the algorithm
How is this a problem? Evolutionary economics said Loss aversion is natural
Something being natural alone does not entail it is not a problem. As an extreme example, cancer is natural. Is cancer not a problem?
Life is noisy
Noice video
It seems that the best and most efficient way to get rid of noise is to remove the human component in decision making.
AI
Which is impossible to do
@@kinesissado9636 No it Isn't. Why do you think it is?
Maybe it will reduce noise, yes, but it will introduce biases
HAHAHAHAHA fake data fake data fake data.