Biggest lesson I learned from my first novella was how differently unfolding game mechanics in narrative works when starting with established characters vs character creation. I learned this approximately 30% in, and it shows ;) Come learn with me here:
https://t.co/bJ0kMJ6fNi
I completely believe Pete. In fact, I have the utmost confidence and faith in all the Tasty personnel hired under Tom and Scott. However, we're in an era where the only thing we've seen is IG layoffs of Tasty personnel ... we haven't seen any IG hires. So sure, I trust Pete, but who else is IG going to hire to transform the company into their vision of Tasty?
How to fix Doctor Who.
Wipe out everything post Peter Capaldi. Most importantly the horrific Timeless Children.
The Dream Lord and The Valeyard are behind the non-events of the 13th, 14th, 15th, and stupid Fugitive Doctor.
Better yet, just never mention them.
Cast Hugh Laurie as the Doctor with a smoking hot actual female companion.
You're welcome.
How to fix Doctor Who.
Wipe out everything post Peter Capaldi. Most importantly the horrific Timeless Children.
The Dream Lord and The Valeyard are behind the non-events of the 13th, 14th, 15th, and stupid Fugitive Doctor.
Better yet, just never mention them.
Cast Hugh Laurie as the Doctor with a smoking hot actual female companion.
You're welcome.
𝐆𝐎𝐎𝐆𝐋𝐄 𝐅𝐈𝐑𝐄𝐃 𝐓𝐈𝐌𝐍𝐈𝐓 𝐆𝐄𝐁𝐑𝐔 𝐎𝐕𝐄𝐑 𝐀 𝐏𝐀𝐏𝐄𝐑 𝐓𝐇𝐀𝐓 𝐏𝐑𝐄𝐃𝐈𝐂𝐓𝐄𝐃 𝐄𝐕𝐄𝐑𝐘 𝐀𝐈 𝐅𝐀𝐈𝐋𝐔𝐑𝐄 𝐖𝐄 𝐍𝐎𝐖 𝐇𝐀𝐕𝐄
In December 2020, Google fired 𝐃𝐫. 𝐓𝐢𝐦𝐧𝐢𝐭 𝐆𝐞𝐛𝐫𝐮, the co-lead of its Ethical AI team, because she refused to retract a 14-page paper called 𝘖𝘯 𝘵𝘩𝘦 𝘋𝘢𝘯𝘨𝘦𝘳𝘴 𝘰𝘧 𝘚𝘵𝘰𝘤𝘩𝘢𝘴𝘵𝘪𝘤 𝘗𝘢𝘳𝘳𝘰𝘵𝘴. The paper had not even been published yet. 𝟐,𝟔𝟗𝟓 𝐨𝐟 𝐡𝐞𝐫 𝐨𝐰𝐧 𝐜𝐨𝐥𝐥𝐞𝐚𝐠𝐮𝐞𝐬 𝐬𝐢𝐠𝐧𝐞𝐝 𝐚𝐧 𝐨𝐩𝐞𝐧 𝐥𝐞𝐭𝐭𝐞𝐫 saying the company's version of the story — that she had resigned — was a lie.
Five years later, every prediction inside that paper has come true.
Hallucinations. The paper argued that large language models trained on internet-scale scrapes would produce systems that sounded fluent but had no actual understanding — statistical parrots that would fool both users and developers into trusting outputs that were structurally incapable of being reliable. The hallucination problem now defines the industry.
Bias amplification. The paper predicted hiring tools that would discriminate against women, healthcare triage that would underperform on Black patients, and lending algorithms that would entrench inequality while presenting their decisions as neutral. Every one has now been documented. Amazon's hiring algorithm penalized resumes that mentioned the word “𝘸𝘰𝘮𝘦𝘯”. Apple Card's credit algorithm gave wives credit lines 10x lower than their husbands at the same financial profile.
Environmental cost. In 2024, Google's emissions were up 48% from 2019; Microsoft's up 29%. Both companies blamed AI infrastructure and quietly walked away from climate commitments they were celebrating the year Gebru was fired.
Undocumentable datasets. The paper said the training data was already too large for anyone to audit. In 2023, researchers found that LAION-5B — used to train Stable Diffusion — contained thousands of images of child sexual abuse material. The companies training on it had no way of knowing.
Model collapse. The paper predicted that AI-generated content fed back into training would degrade languages and centralize cultural power in the small number of firms that could afford to train. A 2024 study found that 57% of new English-language web content is now AI-generated or AI-assisted.
The deeper point of her work is the one the industry refuses to quote. 𝐓𝐡𝐞 𝐭𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐲 𝐰𝐚𝐬 𝐛𝐞𝐢𝐧𝐠 𝐛𝐮𝐢𝐥𝐭 𝐛𝐲 𝐚 𝐬𝐦𝐚𝐥𝐥 𝐠𝐫𝐨𝐮𝐩 𝐨𝐟 𝐫𝐞𝐬𝐞𝐚𝐫𝐜𝐡𝐞𝐫𝐬, 𝐚𝐭 𝐚 𝐬𝐦𝐚𝐥𝐥 𝐧𝐮𝐦𝐛𝐞𝐫 𝐨𝐟 𝐜𝐨𝐦𝐩𝐚𝐧𝐢𝐞𝐬, 𝐫𝐞𝐰𝐚𝐫𝐝𝐞𝐝 𝐟𝐨𝐫 𝐬𝐡𝐢𝐩𝐩𝐢𝐧𝐠 𝐟𝐚𝐬𝐭. Anyone who raised safety or ethics concerns inside the system was ignored, sidelined, or removed. Gebru was making that argument from inside Google. Then Google proved her right by removing her.
Margaret Mitchell, the other co-lead of Ethical AI, was fired two months later for searching her own email for evidence of how Gebru had been treated. The Ethical AI team was dismantled in 90 days.
Gebru founded 𝐃𝐀𝐈𝐑 (𝐭𝐡𝐞 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐞𝐝 𝐀𝐈 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐈𝐧𝐬𝐭𝐢𝐭𝐮𝐭𝐞) in 2021 to do AI research outside the control of the companies that have a financial interest in not hearing the answers. The number of researchers willing to speak publicly about safety inside the major labs collapsed after her firing and has not recovered.
𝐆𝐨𝐨𝐠𝐥𝐞 𝐟𝐢𝐫𝐞𝐝 𝐡𝐞𝐫 𝐟𝐨𝐫 𝐭𝐞𝐥𝐥𝐢𝐧𝐠 𝐭𝐡𝐞 𝐭𝐫𝐮𝐭𝐡 𝐚𝐛𝐨𝐮𝐭 𝐚 𝐩𝐫𝐨𝐝𝐮𝐜𝐭 𝐭𝐡𝐚𝐭 𝐢𝐬 𝐧𝐨𝐰 𝐭𝐡𝐞𝐢𝐫 𝐦𝐚𝐢𝐧 𝐛𝐮𝐬𝐢𝐧𝐞𝐬𝐬. 𝐓𝐡𝐞 𝐩𝐞𝐨𝐩𝐥𝐞 𝐢𝐧𝐬𝐢𝐝𝐞 𝐆𝐨𝐨𝐠𝐥𝐞 𝐰𝐡𝐨 𝐚𝐠𝐫𝐞𝐞 𝐰𝐢𝐭𝐡 𝐡𝐞𝐫 𝐚𝐫𝐞 𝐧𝐨𝐭 𝐚𝐥𝐥𝐨𝐰𝐞𝐝 𝐭𝐨 𝐬𝐚𝐲 𝐬𝐨.
This was absolutely the worst part of learning German, trying to figure out why a church has a vagina, a table has a penis, two tables have a vagina, and a girl is a neutered "it" like a pair of Lego pants.
The Charismatic Voice, a singer-analysis YT channel hosted by an operatic classicist (and giant geek) did a video today on @FiatCelebrity (Jeff Hays') Dungeon Crawler Carl narration!
https://t.co/QSIPuGxnZ1
Two hundred and fifty years ago today — on May 17th, 1776 — the Second Continental Congress called for a day of humiliation, fasting, and prayer.
Today, exactly 250 years later, we gather to do the same. To humbly ask once more for God's mercy and guidance — as we enter the next 250 years of this republic.
And here is the state of the AI witch hunt. As an experiment, a gentleman posts an actual Monet but claims it's AI-generated in Monet's style. The internet proceeds to evicerate the authentic painting, detailing how it could never live up to a real Monet.
Even though our kids won’t see our love “grow old together” from an earthly stand point; they’ll see it from a Heavenly one. And I’ll tell them of our love story any moment I can.
Happy Anniversary to the love of my life.
Top 300 on all of Amazon, is pretty damned for a new book release. Magic & Bullets is #1 in a bunch of New Release categories too.
The only things ahead of me in most categories is stuff like Fourth Wing (Romantasy, which is absurdly huge) and Dungeon Crawler Carl (which is a massive hit sensation). So I will take it!