Luiza Jarovsky, PhD
@luizajarovsky.bsky.social
Co-founder of www.aitechprivacy.com (1,300+ participants). Author of www.luizasnewsletter.com (87,000+ subscribers). Mother of 3.
The AI used by the Utah police to transcribe body camera footage said an officer shapeshifted into a frog.
The reason it did it is even more bizarre...
The state of responsible AI in 2026.
The reason it did it is even more bizarre...
The state of responsible AI in 2026.
January 3, 2026 at 5:23 PM
The AI used by the Utah police to transcribe body camera footage said an officer shapeshifted into a frog.
The reason it did it is even more bizarre...
The state of responsible AI in 2026.
The reason it did it is even more bizarre...
The state of responsible AI in 2026.
Not everything created using AI is slop. Just the vast majority
January 2, 2026 at 2:18 PM
Not everything created using AI is slop. Just the vast majority
Dismissing legitimate AI criticism as 'AI denialism' is inaccurate and ignores the nuances of the debate.
It's also an attempt to paint people offering valid counter-arguments to existing AI trends and policies as stupid.
Don't fall for this trap.
It's also an attempt to paint people offering valid counter-arguments to existing AI trends and policies as stupid.
Don't fall for this trap.
January 2, 2026 at 1:47 PM
Dismissing legitimate AI criticism as 'AI denialism' is inaccurate and ignores the nuances of the debate.
It's also an attempt to paint people offering valid counter-arguments to existing AI trends and policies as stupid.
Don't fall for this trap.
It's also an attempt to paint people offering valid counter-arguments to existing AI trends and policies as stupid.
Don't fall for this trap.
I see no lies here
December 30, 2025 at 1:10 PM
I see no lies here
The 'new internet' in numbers.
We are not prepared for the consequences...
We are not prepared for the consequences...
December 28, 2025 at 1:52 PM
The 'new internet' in numbers.
We are not prepared for the consequences...
We are not prepared for the consequences...
A reminder that in most places, including the United States, AI prompts are NOT copyrightable,
(you cannot claim copyright infringement if someone copies your prompts)
(you cannot claim copyright infringement if someone copies your prompts)
December 28, 2025 at 12:57 PM
A reminder that in most places, including the United States, AI prompts are NOT copyrightable,
(you cannot claim copyright infringement if someone copies your prompts)
(you cannot claim copyright infringement if someone copies your prompts)
Many parents have not realized it yet, but as with drugs and sex, they should discuss AI chatbots use (and what can go wrong) with their teenagers. Times have changed.
December 27, 2025 at 2:45 PM
Many parents have not realized it yet, but as with drugs and sex, they should discuss AI chatbots use (and what can go wrong) with their teenagers. Times have changed.
The current state of education. If we do nothing about it, it's going to get worse.
December 25, 2025 at 3:50 PM
The current state of education. If we do nothing about it, it's going to get worse.
The power of anthropomorphic makeup:
December 24, 2025 at 4:03 PM
The power of anthropomorphic makeup:
After three years of intense hype, the AI honeymoon is officially over.
My final remarks of the year:
My final remarks of the year:
December 23, 2025 at 5:46 PM
After three years of intense hype, the AI honeymoon is officially over.
My final remarks of the year:
My final remarks of the year:
The future of dating.
December 23, 2025 at 2:19 PM
The future of dating.
There is so much automation and machine interactions a person can absorb in their daily routine without compromising wellbeing.
As the AI industry keeps pushing for more than humans can sanely bear, ideas about humans and machines merging gain traction. It's not a coincidence.
As the AI industry keeps pushing for more than humans can sanely bear, ideas about humans and machines merging gain traction. It's not a coincidence.
December 23, 2025 at 10:54 AM
There is so much automation and machine interactions a person can absorb in their daily routine without compromising wellbeing.
As the AI industry keeps pushing for more than humans can sanely bear, ideas about humans and machines merging gain traction. It's not a coincidence.
As the AI industry keeps pushing for more than humans can sanely bear, ideas about humans and machines merging gain traction. It's not a coincidence.
I want my kids to know what AI is, how the training process works, how it can be used, and what can go wrong.
I don't want AI as my kids' teacher, the teaching method, or their main educational tool.
I want AI to be presented as a subject matter, not as a path.
I don't want AI as my kids' teacher, the teaching method, or their main educational tool.
I want AI to be presented as a subject matter, not as a path.
December 21, 2025 at 2:09 PM
I want my kids to know what AI is, how the training process works, how it can be used, and what can go wrong.
I don't want AI as my kids' teacher, the teaching method, or their main educational tool.
I want AI to be presented as a subject matter, not as a path.
I don't want AI as my kids' teacher, the teaching method, or their main educational tool.
I want AI to be presented as a subject matter, not as a path.
Forecast: pressure will continue to grow, and ChatGPT will soon be banned for minors (which would be a positive development, as sycophantic general-purpose AI systems are too risky for vulnerable users).
December 20, 2025 at 2:08 PM
Forecast: pressure will continue to grow, and ChatGPT will soon be banned for minors (which would be a positive development, as sycophantic general-purpose AI systems are too risky for vulnerable users).
Unregulated generative AI is messing up the trustworthiness of scientific research, but what could go wrong
December 20, 2025 at 11:59 AM
Unregulated generative AI is messing up the trustworthiness of scientific research, but what could go wrong
If you look closely, the AI hype makes no sense
December 18, 2025 at 2:35 PM
If you look closely, the AI hype makes no sense
As AI becomes ubiquitous and all-encompassing, active efforts against cognitive deskilling are essential to avoid economic, psychological, and even legal risks.
The 259th edition of my newsletter is out (link below).
Wishing everyone AI with friction in 2026!
The 259th edition of my newsletter is out (link below).
Wishing everyone AI with friction in 2026!
December 17, 2025 at 5:19 PM
As AI becomes ubiquitous and all-encompassing, active efforts against cognitive deskilling are essential to avoid economic, psychological, and even legal risks.
The 259th edition of my newsletter is out (link below).
Wishing everyone AI with friction in 2026!
The 259th edition of my newsletter is out (link below).
Wishing everyone AI with friction in 2026!
The Disney–OpenAI deal marks the end of an era, and I still can’t wrap my head around how casually Disney sold its 100-year-old reputation.
December 14, 2025 at 2:26 PM
The Disney–OpenAI deal marks the end of an era, and I still can’t wrap my head around how casually Disney sold its 100-year-old reputation.
The elephant in the room nobody wants to talk about?
A critical obstacle to broader generative AI adoption (especially enterprise) are AI hallucinations.
However, AI hallucinations do not seem to be fully solvable.
And the investments have already been made.
A critical obstacle to broader generative AI adoption (especially enterprise) are AI hallucinations.
However, AI hallucinations do not seem to be fully solvable.
And the investments have already been made.
December 11, 2025 at 6:53 PM
The elephant in the room nobody wants to talk about?
A critical obstacle to broader generative AI adoption (especially enterprise) are AI hallucinations.
However, AI hallucinations do not seem to be fully solvable.
And the investments have already been made.
A critical obstacle to broader generative AI adoption (especially enterprise) are AI hallucinations.
However, AI hallucinations do not seem to be fully solvable.
And the investments have already been made.
And the future of the entertainment industry is...
...licensing their content to AI companies.
...licensing their content to AI companies.
December 11, 2025 at 3:39 PM
And the future of the entertainment industry is...
...licensing their content to AI companies.
...licensing their content to AI companies.
Most people and companies have not realized it yet, but there is a reason why AI-generated ads are widely hated.
Hint: it has nothing to do with the surrealism or the glitches.
I wrote about it today:
Hint: it has nothing to do with the surrealism or the glitches.
I wrote about it today:
December 10, 2025 at 3:25 PM
Most people and companies have not realized it yet, but there is a reason why AI-generated ads are widely hated.
Hint: it has nothing to do with the surrealism or the glitches.
I wrote about it today:
Hint: it has nothing to do with the surrealism or the glitches.
I wrote about it today:
Listen to Nicolas
December 8, 2025 at 7:46 PM
Listen to Nicolas
AI regulation does NOT mean an AI ban.
December 8, 2025 at 4:47 PM
AI regulation does NOT mean an AI ban.