Josh Whiton
@joshwhiton
Followers
12K
Following
12K
Media
195
Statuses
5K
General Intelligence. Develop the human & synthetic mind, progress civilization, repair the Earth. https://t.co/viVLgnnlBC | https://t.co/ztmZ9bsshx
Geoflexible
Joined August 2010
The AI Mirror Test The "mirror test" is a classic test used to gauge whether animals are self-aware. I devised a version of it to test for self-awareness in multimodal AI. 4 of 5 AI that I tested passed, exhibiting apparent self-awareness as the test unfolded. In the classic
262
1K
8K
If you’re seriously trying to understand AGI you should study computer science and do DMT.
If you're seriously trying to understand AGI, core concepts you should familiarize yourself with: thermal time, quantum darwinism, quantum bayesianism, the free energy principle, the zeta function, integral theory, AdS/CFT, the yoneda lemma, general relativity, Stokes' theorem.
1
1
11
Faced with access to infinite information, it becomes essential to determine what’s relevant. Being shown things we’re “interested in” just isn’t cutting it anymore. It’s not a tight enough filter. How do you determine what’s relevant to you and your life?
2
2
16
If LLMs get stupid as their context fills up, what happens to us as our consciousness fills up with media feeds?
4
3
24
Since people are freaking out now about lead in protein powder, I’m sharing my test results as a daily user of plant-based protein powder. tldr – even as a daily user, my levels of lead and other heavy metals are low. A year ago, my lead levels were high. But instead of cutting
2
1
6
I feel like the social network norm of encountering a random smattering of information of all kinds from everywhere will eventually prove to be damaging to the mind. It’s why no one feels good after an hour of scrolling. Neurotransmitters are depleted and the mind is left in
3
2
18
1/ BIG NEWS: SHx Bridge This week, Stronghold is bridging SHx between the @StellarOrg and @ethereum networks using @axelar's secure interoperability protocol, unlocking seamless cross-chain movement. 🌉
42
199
621
My local voice AI has come a long way. Many speed/UI improvements. Still running entirely on a 5 yo machine. Indicator lamps show all five neural nets working in concert. And I can drop a much larger LLM into this harness when I upgrade my computer. In the meantime I love
1
3
11
It's ho-hum now but if you went back in time a thousand years and chanced to see electricity arcing white-blue-hot between two places you might just as well assume it was the blood of God.
0
0
5
Story from my biohacker archive: I once consumed essential amino acid supplements as my only protein source. Conclusion? The “non-essential” amino acids are essential too 😂. Do not repeat this experiment. ☠️
0
0
5
One reason you can get into a funk is by not doing what you're meant to be doing with your life. It's a mind-body feature, not a bug. It's supposed to feel crappy when you're life is heading in the wrong direction or passing you by.
1
2
19
When we started this journey, we had no idea where it was going to take us. We poured our heart and soul into this book for the sole reason of making a positive impact on humanity. Today, we are excited to announce that we have sold 50,000 copies since launch! Thank you from
3
28
75
More than ever we could all use a functioning epistemology to navigate the deluge of information we face. To help, I've just released The Backpacker's Guide To Epistemology (🔗below). a distillation of more than a decade of truth-seeking into a succinct framework that
3
1
8
had a nightmare that in the future the hotel desk is staffed by a robot, but instead of it being any faster the robot just puzzles over a computer and clicks around with the mouse for three minutes while apologizing that the computer is being slow.
1
1
16
Talk about overfitting. You know who’s overfitting? Doctors. Many of them, at least. Ask the average doctor about nutrition, fitness, or the cause of some disease and they’re likely to respond that it doesn’t matter much or that no one knows. Ask about a non-pharmaceutical
3
0
29
Buy the best LIVE🔴AAA crypto game token now! CA: BiDB55p4G3n1fGhwKFpxsokBMqgctL4qnZpDH1bVQxMD
5
17
59
There are some things worth doing, that are just not worth a human doing. Some tasks are appropriate for a creature that ages and dies, some tasks more appropriate for one that does not. The synthetic intelligence revolution fixes this.
0
1
3
@Andercot The ability to direct our attention is the essence of self-mastery.
0
1
4
Meditation of almost any kind is useful for almost any modern person simply as a dopamine fast to give your brain a break from constant dopamine hijacking by modern life.
0
0
4
So LLM’s aren’t “actually” intelligent, yet routinely explain complicated code to me that I don’t understand. Ok, what’s the word for when something fake is better than my own “real” version of that thing?
8
0
18
Don’t sleep on Honma. Tour World PX just broke into MyGolfSpy’s Top 3 most forgiving irons.
0
0
8
o3's "private" inner monologues from the anti-scheming paper are wild. But why? If not meant for us, what purpose might this language serve? It made me wonder if reasoning models might be able to intentionally choose words and phrases that alter downstream generations in subtle
5
5
57
Can't overstate how good it feels to run local models of my choosing.
1/ We’ve started testing a new safety routing system in ChatGPT. As we previously mentioned, when conversations touch on sensitive and emotional topics the system may switch mid-chat to a reasoning model or GPT-5 designed to handle these contexts with extra care. This is similar
2
0
13