Thanks to @jdlahart.bsky.social at @wsj.com for chatting with me about the tough job market for youngsters. In the end, a lot of the problem is just a big cyclical one: hiring is weak. That hurts people looking for jobs, who tend to be younger than average.
www.wsj.com/economy/jobs...
My experience is a reasonable reference. I make correct assessments of the world. I am good. My group is a reasonable reference. My group is good. journals.sagepub.com/doi/10.1177/...
not all lines going down
that's a good one
My friend tells me the idiom "crossed the Rubicon" is overused. What's are the alternatives?
Great postdoc opportunity here academicjobsonline.org/ajo/program/...
Study the role of worker voice in AI-driven organizational change with the excellent scholars at Cornell's ILR School.
I appreciated speaking with @talsmith.bsky.social for his excellent analysis on the question of data vs sentiment in the U.S. economy www.nytimes.com/2025/03/31/b...
For Equal Pay Day 2025, we looked at Glassdoor ratings & found women rate their employee satisfaction 3.5% lower than men. While this gap is present in satisfaction w/ compensation & benefits (-3.6%), other factors like senior mgmt (-3.4%) and career opportunities (-3%) also show a gap as well.
I think it's important that AI tools get stress-tested and held to account for bad info. Need to push back on the tendency for people to treat them as magical oracles. But believing the opposite, that most genAI is just constantly wrong, is a real misperception.
If you look at this Paris Marx tweet quoted above and the responses to it, you see a lot of people confidently asserting that generative AI is factually incorrect most of the time. I've found this to be a pretty pervasive belief in certain quarters.
Just heard someone cite this stat as something to the effect of "ChatGPT is wrong 60% of the time" which is wrong and not what the CJR study found. CJR found that AI search tools got *news citations* wrong that often. That's bad, but it's not the same as genAI getting facts wrong generally.
I like this addition to the abundance discourse from @dsquareddigest.bsky.social. It's only natural that things take longer to build over time (how much longer is a different question) backofmind.substack.com/p/cybernetic...
NEW: The Census Bureau hasn't been recruiting and retaining enough interviewers for three key surveys, including the source of the monthly job report, and lacks a plan for building up a field representative staff, the bureau's internal watchdog, Commerce OIG, finds
www.oig.doc.gov/wp-content/O...
McDonald's is looking to implement a “generative AI virtual manager” www.wsj.com/articles/mcd...
Last thing. BLS does amazing work to create timely, accurate info about America's working families & employers, a huge public good. They are there for us & we need to show up for them.
If you are a labor researcher or care about employment, follow & join Friends of BLS.
www.friendsofbls.org/join
Really odd trend over time. Falling pre-GFC but rising into 2020, big COVID dip then back up. Are we just resuming a secular post-2010 trend? Have lending standards gradually shifted over time (and not just with biz cycles)?
Sloan Foundation announced a program to support postdocs in philosophy, sociology, and metascience who want to understand how AI will affect science.
Pass it along to anyone who might be interested, the Institute for Complex Social Dynamics might be interested in hosting
sloan.org/programs/dig...
40% of students who start college don't finish. Some high schools that once pushed "college for all" are grappling with that reality, and providing teens with more career-focused learning.
www.nytimes.com/2025/03/06/u...
Cool new opportunity @equitablegrowth.bsky.social
#FundSocSci #EconSky #Sociology #PolicySky #EduSky
equitablegrowth.applicantpro.com/jobs/3632377
www.washingtonpost.com/politics/202...
Not a good trend. Wonder if this is also showing up in IRS admin data www.wsj.com/personal-fin...
Curious what you make of scholarly survey work like this looking at broad usage patterns. This has ~10% of U.S. respondents using genAI tools every day at work as of late 2024. www.nber.org/papers/w32966
List of CDC information that appears to have been taken down so far:
1. CDC Atlas
2. CDC Youth Risk Behavior Surveillance System
3. HRSA Target
4. Language removed for LGBT work
5. CDC Social Vulnerability Index
It’s official: all funding requests made on Tuesday by postdocs for their salaries have been unilaterally rejected by NSF.
Parental support for permitting their children "to misrepresent something about their identity … to qualify for additional resources or initiatives (e.g., scholarships, affirmative action).” Goes up with income, but only among those who believe in upward mobility! papers.ssrn.com/sol3/papers....
Fun Baffler piece about AI advances in HR tech. This is a new one for me
thebaffler.com/outbursts/hi...
Featured in the latest Bulletin on Retirment and Disability: "Health Inequality and Economic Disparities by Race, Ethnicity, and Gender"
https://www.nber.org/brd/20244/health-inequality-and-economic-disparities-race-ethnicity-and-gender
They don’t name em like they used to
They discuss some possible mechanisms here, short-staffing among them. They claim their results are consistent with work intensification leading to injuries but I don't think they can say whether that's due to short-staffing or employers trying to get more bang for their buck (so to speak)
