ARTexplains, Ph.D.

Everything is interesting! ARTexplains Science & History on YouTube. My book is called "Incredible Consequences of Brain Injury." Ph.D. in Cognitive Psychology.

[ARTexplains, Ph.D. @ARTexplains]

My book "Incredible Consequences of Brain Injury: The Ways your Brain can Break" releases on November 11! Pre-orders are open via Amazon, Bookshop, and several other places! Learn more here: https://t.co/RPwUYY9vbk

[ARTexplains, Ph.D. @ARTexplains]

https://t.co/PoS7w284dO

[ARTexplains, Ph.D. @ARTexplains]

@Josh_Isnt Aww I didn't get to finish it. Will it go back up soon?

[ARTexplains, Ph.D. @ARTexplains]

(*￣∇￣)ノ

[🐢🐢🐢John Green🐢🐢🐢 @johngreen]

I HAVE RETURNED TO MAIN because I am FURIOUS that @JNJGlobalHealth is preventing millions from accessing life-saving treatment with bedaquiline, thereby costing innumerable lives and increasing the global risk of multidrug-resistant tuberculosis. https://t.co/AbTdNzC1Sm

[🐢🐢🐢John Green🐢🐢🐢 @johngreen]

One of the reasons tuberculosis continues to infect and kill more people is because you refuse to allow the widespread distribution of affordable bedaquiline, which is essential to stopping the spread of multidrug-resistant tuberculosis and curing those infected by it.
It takes some gall to wonder aloud what the problem is when you know that you are the problem. You've had 20 years of a bedaquiline monopoly. You've made your money. ENOUGH.

[ARTexplains, Ph.D. @ARTexplains]

@sciam This article is riddled with typos; give it another read-over before publishing

[ARTexplains, Ph.D. @ARTexplains]

@ESYudkowsky My biggest note to you is to make your tweets more readable by a general audience -- and I think you're getting that feeling too. Artificial Intelligence Alignment is important for the general public, but we aren't communicating it well enough to that general public.

[ARTexplains, Ph.D. @ARTexplains]

Can't we just go back to the normal way where we google our symptoms and then assume we have brain cancer?

[ARTexplains, Ph.D. @ARTexplains]

Therapists: AI psychotherapy won't work; machines can't have empathy. They can't relate to your experiences by putting themselves in your shoes

OpenAI: The new GPT wears shoes

Therapists:

[ARTexplains, Ph.D. @ARTexplains]

Turns out wearing shoes was an emergent property of large language models 🤷🏻‍♂️ But can they overprescribe antidepressants? Or is psychiatry safe?

[ARTexplains, Ph.D. @ARTexplains]

@MatthewGaydos @phildunster Wait, what was the TFIOS reference?

[The Onion @TheOnion]

More Realistic Meat Substitute Made From Soy Raised In Brutally Cruel Conditions https://t.co/4cOABb1DM4

[ARTexplains, Ph.D. @ARTexplains]

Finally increased the size of my cursor because I kept losing track of it. Also made it purple, as a treat. Do what you can to help your eyeballs, friends!

[ARTexplains, Ph.D. @ARTexplains]

New video! Why do some people believe that they are dead? Cotard Syndrome, or Walking Corpse Syndrome, is the answer. https://t.co/xUmF8ipUpH

[ARTexplains, Ph.D. @ARTexplains]

@josegeeklml @doomquasar > attempt to ford the river
> caulk wagon and float it across

[ARTexplains, Ph.D. @ARTexplains]

"It’s fascinating... to read doomsday predictions, but it’s even more fascinating to watch what happens to the reasoning of true believers when the prediction flops..." -Tavris & Aronson: Mistakes Were Made (But Not By Me)

Will #AGI progress create "flop" doomsday predictions?

[ARTexplains, Ph.D. @ARTexplains]

Gonna need a USB breathalyzer to activate this, so that when your users are drunk, they don't mass email an itemized list of how horny they are.

And then figure out how to safeguard the actions of people who are tired, angry, high, addicted, logged in on a public computer...

[Matt Shumer @mattshumer_]

Using Personal Assistant is incredibly intuitive and straightforward.

Just describe what you want it to do, and it will automatically operate Chrome for you to achieve your task.

[ARTexplains, Ph.D. @ARTexplains]

Awful development. These are all features that increase the existential risk of AI -- they make it more difficult for humans to intervene if an AI agent's goal is misaligned with human goals. We need more safeguards, not fewer.

[Nathan Lands @NathanLands]

The main features of AutoGPTs:
1) assign tasks/goals to be worked on automatically until completed
2) chain together multiple GPT-4s to collaborate on the tasks
3) internet access and ability to read/write files
4) memory to know what has been done

[ARTexplains, Ph.D. @ARTexplains]

I bet the local gastropub just calls it Dew-Kissed Herba so they can charge you more

[ARTexplains, Ph.D. @ARTexplains]

Maybe a nice royal purple?

[ARTexplains, Ph.D. @ARTexplains]

Considering organizing my bookshelves by color. Any books from violet to magenta that y'all'd recommend?