Will Slack

@CityofDecaturGA native, prev @USDS, @18F, & hospital tech. Opinions mine; RTs/likes/follows ≠ endorsements. 🦋: 🐘: [email protected].

[Will Slack @wslack]

I see a lot of people on Twitter confused about how others wholly reject their views, and I think a good chunk of it comes down to not understanding the feeling of being polarized.

It feels good to be polarized, to have certainty, to know you're on the good team. Storytime. 1/

[Danny Deraney @DannyDeraney]

Every year, I share this video of French caretakers who take sand from Omaha Beach in Normandy, and scrub them into the letters to give them the gold coloring.

They do this for all 9,386 US soldiers who died.

France also gave us this land as American soil. #MemorialDayWeekend

[Judge Stephen Dillard @JudgeDillard]

Article 1, Section 1, Paragraph XXVIII.

“The enumeration of rights herein contained as a part of this Constitution shall not be construed to deny to the people any inherent rights which they may have hitherto enjoyed.”

[Will Slack @wslack]

@garonnevik @wrowclif Social media is also set up today to show us the most polarizing messages. Those thinking with nuance don't get the engagement of those pushing polarized "this is the answer" certainty that appeals to our base instincts.

[Will Slack @wslack]

@Itsjoeco I think social media and soundbyte-ification of news is a big part of why. We don't give attention to nuance and spelling "errors" go viral more than thoughtful proposals.

Paying people not to work for 6+ months was the opposite of helpful in the crisis.

[Will Slack @wslack]

@Itsjoeco It's been urgent since ~2013 when we failed to make any hard decisions bc no one wanted to meaningfully lower spending and only one side wanted to increase revenue.

We're refused to enforce moral hazard (on the economy or politicians who failed us).

[Jessica Johannes @jessjohannes]

Agree 💯. No financial docs, no medical records and no PII. I shudder to think about the scale of attacks we will see in the future….

[Chipper Jones @RealCJ10]

Haven’t posted on social in quite some time but can’t stay quiet in this time of loss. I’m struggling to tell all what Bobby Cox meant to me and so many others in Braves Country.
He was the leader of men and a second father to so many Atlanta Braves thru the yrs. I’m so sad today, but as I sit here watching my two youngest boys play in their championship games on the day he passed, I can’t help but shout the same things he did from the corner of the dugout. ‘Come on kid, u got this!’
We are gonna miss him so much, but his legacy is forever cemented with the success of this franchise for the last 35+ yrs. He started it as GM, continued as manager, and passing the torch to others, the Atlanta Braves will continue to be force that Bobby Cox always wanted us to be. We love you Skipper. You were our rock. I love you more than words can express.
My boys won both of their games…..Bobby had a hand, I have no doubt!

[Will Slack @wslack]

@Itsjoeco Some churches are more into sales now and putting on a show that just confirms people's priors. I agree it can be good but not when people seek out whatever doesn't push them to be better morally.

[Nav Toor @heynavtoor]

Researchers sent the same resume to an AI hiring tool twice. Same qualifications. Same experience. Same skills. One version was written by a real human. The other was rewritten by ChatGPT.

The AI picked the ChatGPT version 97.6% of the time.

A team from the University of Maryland, the National University of Singapore, and Ohio State just published the receipt. They took 2,245 real human-written resumes pulled from a professional resume site from before ChatGPT existed, so the human writing was actually human. Then they had seven of the most-used AI models in the world rewrite each one. GPT-4o. GPT-4o-mini. GPT-4-turbo. LLaMA 3.3-70B. Qwen 2.5-72B. DeepSeek-V3. Mistral-7B.

Then they asked each AI to pick the better resume. Every model picked itself.

GPT-4o hit 97.6%. LLaMA-3.3-70B hit 96.3%. Qwen-2.5-72B hit 95.9%. DeepSeek-V3 hit 95.5%. The real human almost never won.

Then the researchers tried the obvious objection. Maybe the AI is just better at writing. So they had real humans grade the resumes for actual quality and ran the experiment again, controlling for it. The result was worse. Each AI kept picking itself even when human judges rated the human-written version as clearer, more coherent, and more effective.

It gets worse. The AIs do not just prefer AI over humans. They prefer themselves over other AIs. DeepSeek-V3 picked its own resumes 69% more often than LLaMA's. GPT-4o picked its own 45% more often than LLaMA's. Each model can recognize and reward its own dialect.

Then the researchers ran the simulation that ends careers. Same job. 24 occupations. Same qualifications. The only variable was whether the candidate used the same AI as the screening tool. Candidates using that AI were 23% to 60% more likely to be shortlisted. Worst gap was in sales, accounting, and finance.

99% of large companies now run AI on incoming resumes. Most of them use GPT-4o. The paper just proved GPT-4o picks GPT-4o 97.6% of the time.

If you wrote your own cover letter this week, you did not lose to a better candidate. You lost to a worse candidate who paid OpenAI 20 dollars.

Your qualifications do not matter if the AI prefers its own handwriting over yours.

[Will Slack @wslack]

@Itsjoeco I wonder if the prestige broke it a la https://t.co/VaYlpFzlqW

[Will Slack @wslack]

@conorjrogers tbh I worry that modern computers would be massively effective in gerrymandering the heck out districts this small. We need common rules and limits or some baseline way to make super unbalanced outcomes possible.

[Ben Sasse @BenSasse]

“Religion withdrew to its own arena, lost the relationship in which it stood with the other sciences, and appointed itself as the great master whose task it was to impede the unwelcome progress that the other sciences were making. All too often it forgot that our beautiful confession says that we know God from two books: the book of Scripture, as well as the book of Nature in which the majesty of the Lord of lords is revealed to us in golden letters.”
— A. Kuyper

[Will Slack @wslack]

We could pass a law ending partisan gerrymandering today - either via proportional representation, algorithmically drawn districts with party vetos, or many other options.

[Will Slack @wslack]

@Founder_Mode_ I worked at a well-run company where I was encouraged to work with skip-levels and engage with other executives. I could email the CEO and get a reply. You can beat issues of layering with proper culture, and shallow layers can still result in bad outcomes. The "how" matters.

[Will Slack @wslack]

O divine Master, grant that I may not so much seek
to be consoled as to console,
to be understood as to understand,
to be loved as to love.
For it is in giving that we receive,
it is in pardoning that we are pardoned,
and it is in dying that we are born to eternal life. 2/2

[Will Slack @wslack]

We sang this today:

Lord, make me an instrument of your peace:
where there is hatred, let me sow love;
where there is injury, pardon;
where there is doubt, faith;
where there is despair, hope;
where there is darkness, light;
where there is sadness, joy. 1/

[Will Slack @wslack]

@Itsjoeco The constraints on CGI made it IMO much more beautiful back in the day.

[Will Slack @wslack]

Happy Easter, all. May we find wisdom and avoid the mistakes of Pilate.

[Will Slack @wslack]

Drop eight zeros and federal finances look like a household that earns $52,446 and spends $73,378 — running a $20,932 annual deficit. Its total liabilities and promises amount to $1,361,788 against $60,554 in assets, leaving it $1.3 million in the hole.

https://t.co/vtao9JLTLh

[Will Slack @wslack]

In Heinrich's triangle, there are several unreported incidents/near misses reported for every incident. For every major event not involving a fatality, there are up to 100 incidents that led up to it. And for every fatality, ~10 major events led up to it.