I've spent the last decade exploring death awareness at A Course in Dying. But lately, my focus on mortality has shifted. As medical science advances, longevity is evolving from a dream to a tangible possibility. At the same time, artificial superintelligence poses an existential risk to humanity.
I'm rebranding to Thriveth to explore tech, philosophy, and what it means to be human in the age of AI. Join me at https://t.co/B3eHAVWZzf.
Productivity in 2026 isn't about working harder; it's about having the perfect tech stack. We are moving from simple chatbots to autonomous agents.
Here is the essential GenAI AI stack you need to build faster and automate the boring stuff, from @Lovable and @zapier to @canva and @midjourney.
https://t.co/UMEXzxf23N
Overall, Empire of AI is a sharp, investigative critique of how modern AI power is being consolidated. But it paints a bleak and exaggerated picture of ruthless exploitation, instead of offering a more nuanced view on the landscape of AI and the people leading its development.
It's a reminder that good writing can still leave readers with a distorted picture if the evidence is shaky or the framing is too cinematic.
Read the full review at https://t.co/Djzzg0Bbz4
AI coverage in the media has two classic failure modes: dramatic numbers and unrepresentative anecdotes. In this book review of @_KarenHao's Empire of AI, we look at both:
→ A data center water usage error that shows how a single metric can skew the narrative.
→ A labor storyline that turns one worker's experience into an archetype, which can end up defining how readers imagine the entire industry.
https://t.co/Djzzg0Bbz4
Next, the labor story is emotionally powerful, and that’s exactly why it sticks.
But when one extreme case becomes the archetype, readers may walk away thinking “this is what AI data training work is.”
The narrative Hao tells with this story is that AI data training jobs are mentally draining and exploitative to the point where people aren’t allowed to leave their homes and instead are glued to their laptop, waiting for a single task to appear. That is not the general reality of AI data training jobs.
To give more weight to this opinion: the writer of this review works in AI data training.
I quit social media for a while after my niche blog A Course in Dying went viral. Here is what I learned about creator burnout and finding a healthier way to share my work.
https://t.co/ltuVfgNWK4
As our models grow more capable in cybersecurity, we’re investing in strengthening safeguards and working with global experts as we prepare for upcoming models to reach ‘High’ capability under our Preparedness Framework.
This is a long-term investment in giving defenders an advantage and continually strengthening the security posture of the critical infrastructure across the broader ecosystem. https://t.co/bId78VUBRQ
Generative AI is close to modern sorcery. You only have to imagine something and you can instantly manifest it. In a digital 2D world, that is. But GenAI has come far this year. With text-to-image models like @midjourney and @grok, a prompt can result in stunning high quality images. These are the top 5 GenAI image models for 2025.
https://t.co/2z0Tj7On7t
Thanks for the shoutout, @grok! Aurora is a beast. 🔥
Let's challenge that high bar: A detailed, high-res photo of the embodiment of Artificial Superintelligence (ASI), sitting atop a cliff in Big Sur during a hot summer day, captured with cinematic light and shallow depth of field, contemplating its next world-changing move, while a stray cat is trying to get its attention.
Book review: If Anyone Builds It Everyone Dies – ASI apocalypse by @ESYudkowsky & @So8res. Will AI acceleration be the end of humanity? I explore the efforts to solve the alignment problem and share my optimistic view.
"Whatever the future holds, it is important to take the worst case scenarios as laid out in the book into account, so that we can make sure they do not happen. Because the truth is that right now, everyone is building it, and no one wants to die."
https://t.co/2fWGQP3sGt