Florian Keppeler

It is great to work with this awesome team! 👐

How one college professor uses AI in the classroom Harvard Kennedy School Professor Teddy Svoronos is mostly optimistic on the benefits of AI in higher education.

Super interesting interview on the use of AI in the classroom by the expert on AI and pedagogy at HKS @tedsvo.bsky.social - he's right, this is the big AI question we need to talk about more. www.wbur.org/upnext/2025/...

*who

@mkfeeney.bsky.social - people how can teach #beagle can teach everyone 🤪

Hvis du har noget at spise, så fungerer det med kommandoer med enhver beagle… 🍪

Flot tysk!!! 🥰

YouTube video by Peter Hald Feeding Wilma #dog #beagle

#WilmasVejledning i veludført vovsning:

"Hvad sker der, hvis en uimodståelig kraft møder et urokkeligt objekt?"

Beagler er berømte for opdragelsesresistens, men Wilma er opdraget af en tysker ( @flokeppeler.bsky.social ).

Tysk er et godt sprog til kommandoer!

youtube.com/shorts/rbG1O...

🔗 doi.org/10.1111/puar...

Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk

End🧵

➡️ Whether focusing on policy, technology, or management, fellowship programs help to place specialized talent across federal, state, and local government roles. 🏛️

💡Fellowship programs in the public sector have been successful in improving recruitment/retention of early career talents.

4/🧵

➡️ In contrast to traditional public hiring, fellowships offer a more accessible and streamlined path into government, clear timelines, support, and interesting tasks. 📋

➡️ Fellows are over 30 percentage points more likely to work in government, and this effect lasts about eight years. 📈

3/🧵

..fellowship programs make it possible to do so without entering uncharted waters. 🐕

The @thepeoplelab.bsky.social researchers reveal how these programs are facilitating public sector recruitment. The @pareview.bsky.social article followed over 2,000 fellows across 19 years and found:

2/🧵

#Wilmasreview
Wilma 🐶 may be hesitant about making the next step, like people who people who want to take their next career step, maybe into the public sector. New research by Sciepura, Wall and @elizabethlinos.bsky.social shows that...

1/🧵

...investing in collaborative human-AI diagnostic systems could significantly reduce costly errors.

🔗https://doi.org/10.1073/pnas.2426153122

Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk

End 🧵

➡️ This performance advantage held across different medical specialties and experience levels, suggesting that the benefits of human-AI collaboration may not be limited to specific areas of medicine.

💡 For policymakers: Rather than viewing AI as a replacement for physicians, ...

3/🧵

➡️ When a doctor and a computer program work together to diagnose a patient, they can make a better diagnosis than if they worked alone! 🏥
➡️ Humans and AI-tools make complementary errors. This means that diagnostic errors can be reduced through collectives of doctors and AI tools! 📊

2/🧵

#Wilmasreview
Wilma 🐶 has been thinking about adding AI-teammates to her pack. Would that help her find more delicious treats?

Well, this might work for making complicated decisions, like medical diagnoses. A study by Zöller et al. shows:

1/🧵

Taxing reproduction: the full transfer cost of rearing children in Europe | Royal Society Open Science What are the intergenerational resource transfer contributions of parents and non-parents in Europe? Using National Transfer Accounts and National Time Transfer Accounts for 12 countries around 2010, ...

🔗 doi.org/10.1098/rsos...

Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk

End🧵

💡If Europe wants to address its fertility challenges and aging population, it may want to support parents - they contribute more resources while receiving less recognition for their societal investment.

4/🧵

➡️ Even Nordic countries like Sweden and Finland, known for family-friendly policies, don’t reduce this implicit tax burden on parents. These policies help parents work, but don’t lower the total cost parents pay to raise the next generation of taxpayers. 🇸🇪🇫🇮

3🧵

➡️ When you count ALL transfers (taxes, money, and time), parents contribute 2.66 times more to society than non-parents. 💰⏰

➡️ While non-parents primarily contribute through taxes, parents provide additional private transfers - mothers mainly through time, fathers mainly through money. 😰

2/🧵

#Wilmasreview

Flashback: Wilma 7 years ago as a puppy 🐶 It takes a lot of effort to raise a puppy, but a human baby takes extra effort! 👶

A new study by Pieter Vanhuysse, Márton Medgyesi and Róbert Iván Gál shows:
1/🧵

➡️ The effect is remarkably consistent across contexts. Whether applying for jobs, seeking college admission, or participating in research studies, the same pattern emerges. 🔬

🔗 doi.org/10.1073/pnas...

Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk @pnas.org

End 🧵

➡️ People act more analytical under AI assessment. When participants knew an AI would evaluate them, they reported approaching tasks more analytically. 🤖
➡️ This behavior change stems from beliefs about AI priorities. People believe AI systems prioritize analytical characteristics. 📊

2/🧵

#Wilmasreview
Wilma 🐶 tried in vain to get through this hole. She probably analyzed it wrong. Perhaps it would have been better if this had been done under AI assessment. Why?

A study by Goergen, de Bellis, and Klesse shows that...

1/🧵

🔗 doi.org/10.1073/pnas...

Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk

End🧵

💡 So, if people judge whether or not they deserve something based on the outcome, not the effort they put in, we may need to rethink the way we give feedback and the way we incentivize people - especially in situations where performance is heavily influenced by chance.

4/🧵

➡️ Working extremely hard on a difficult task didn't make people feel more deserving of rewards if their results were poor. 🤔

➡️ Most surprisingly, people still felt they deserved rewards even when they knew luck played a major role in making their task easy. 🍀

3/🧵

...some surprising findings. Here's what Wilma learned:

➡️ People felt entitled to rewards based on how well they performed, not how hard they worked. Even participants who completed effortless tasks but achieved good results felt they deserved payment! 😮

2/🧵

#Wilmasreview

Wilma 🐶 sits proudly next to a child's beautiful drawing of her, and while she put zero effort into posing, she certainly considers this masterpiece quite an achievement worth an extra treat! 🤩

Speaking of effort versus achievement, a study by Cusimano, Kim, and Wong reveals...

1/🧵

🔗 doi.org/10.1016/j.le...

Wilma, PhD (Pretty helpful Dog) and #postdog 🐾
@au.dk

End🧵

💡 No evidence suggests that young leaders are less effective than older ones. Organizations serious about combating skill shortages should examine their leadership selection processes for age bias.

4/🧵