Michael Smith

5.3K posts

Michael Smith banner
Michael Smith

Michael Smith

@Morphenius

San Francisco, CA, USA Katılım Kasım 2012
154 Takip Edilen2.5K Takipçiler
Sabitlenmiş Tweet
Michael Smith
Michael Smith@Morphenius·
Basically no one is taught math. Sure, people get thrust through something called "math class". They learn things called "math". And some learn the true Art DESPITE how it's taught. But most folk never see math. Instead they're taught COMPUTATION.
English
19
70
601
189.4K
Michael Smith
Michael Smith@Morphenius·
@binarybits I miss the letter “thorn”. English’s theta. It just morphed in shape until it looked just like a “y” and then got discarded. (That’s why “ye olde shoppe”: the “ye” had a thorn & would be pronounced “the”.)
English
0
0
1
62
Timothy B. Lee
Timothy B. Lee@binarybits·
People have started abbreviating Anthropic as "Ant" and I don't like it. The th is a single phoneme. You shouldn't split it up like that.
English
54
11
549
19.7K
Michael Smith
Michael Smith@Morphenius·
@robertskmiles @eurydicelives We already have models for how FTL could work. We just can’t build them yet. The apparently firm limitation is that we can’t smoothly accelerate through & beyond c. Have to get FTL effects cleverly.
English
0
0
1
39
eurydice
eurydice@eurydicelives·
tell me something you think a superintelligence will never be capable of doing
English
70
3
102
20.6K
Michael Smith
Michael Smith@Morphenius·
Being polite to AI is good for character. It’s bad form to practice treating a person-like interaction from a place of entitlement. All beings deserve compassion & dignity. It’s a good habit even if it’s silly sometimes (which maybe it never is).
Aakash Gupta@aakashgupta

Sam Altman said people saying “please” and “thank you” to ChatGPT costs OpenAI tens of millions of dollars a year in compute. 67% of Americans do it anyway. Run the math on why. A 2024 Waseda University study tested LLM responses across politeness levels in English, Chinese, and Japanese. Impolite prompts produced measurably worse outputs: more bias, more errors, more refusals. Moderate politeness consistently beat both extremes. The mechanism makes sense once you see it. Polite prompts pattern-match to higher-quality training data. When you write “Could you help me structure this analysis?”, the model pulls from professional, well-reasoned text. When you write “give me the answer,” it pulls from Reddit. Google DeepMind’s Murray Shanahan explained it simply: the model is role-playing a smart intern. Treat the intern like a colleague, you get colleague-quality work. Bark orders, you get minimum-viable compliance. Now look at the cost side. OpenAI handles over a billion queries daily. Each GPT-4 query uses roughly 2.9 watt-hours, ten times a Google search. But OpenAI just raised $40 billion at a $300 billion valuation. Tens of millions in politeness tokens is a rounding error on a rounding error. 67% of users do it anyway, and 55% of them say it’s because it’s “the right thing to do.” They’re maintaining a behavioral habit that governs every other interaction in their life. The parent who teaches their kid to say please to Alexa isn’t doing it for Alexa. They’re doing it because the alternative is raising someone who learns that being rude gets faster results. Telling 900 million people to stop saying thank you so OpenAI can save 0.01% of operating costs is the most engineer-brained optimization take on the internet. You’re training yourself to treat every interaction as a transaction. And that habit doesn’t stay in the chat window.

English
0
0
12
1.7K
Michael Smith
Michael Smith@Morphenius·
There's sth I'm stewing on how to convey to 21st century transhumanists. How it's not an ESCAPE FROM being human, but an EXTENSION of it. That's what "alignment" is really about. You can't have hope if your envisioned future arises purely from reaching for "not the bad thing".
English
0
1
6
286
Michael Smith
Michael Smith@Morphenius·
Deathism doesn't always sound like it's about death. Sometimes it shows up as the "wisdom" of giving up & letting what's sacred decay & die. Don't let death intimidate you. Doesn't matter what mask it wears. You know in your heart what's precious. Cherish & protect that forever.
Michael Smith@Morphenius

@IvanVendrov @gmiller Start by assuming there's a way. Take your ignorance of WHAT that way is as a challenge. Not knowing how to solve a math problem isn't a good reason to let it defeat you. But you can make it one if you want. Popular choice.

English
0
0
5
281
Michael Smith
Michael Smith@Morphenius·
@IvanVendrov @gmiller Start by assuming there's a way. Take your ignorance of WHAT that way is as a challenge. Not knowing how to solve a math problem isn't a good reason to let it defeat you. But you can make it one if you want. Popular choice.
English
0
0
4
349
ivan
ivan@IvanVendrov·
@gmiller stop them how? how do you propose we fight trillions of dollars in economic incentives & now the most powerful militaries in the world? if I had a good plan to slow things down I would. for now it seems the best I can do is try to steer well through the singularity
English
5
0
13
1.4K
ivan
ivan@IvanVendrov·
a mood I'm really missing in the current AI discourse is grief yes things might go terribly and yes we might see glories beyond imagining but no matter what, we will lose much of what it has meant to be human, forever. I'd like to be with that grief more, and held in it.
English
86
49
824
58.3K
Michael Smith
Michael Smith@Morphenius·
@IvanVendrov You’re more likely to be right if you encourage yourself and everyone around you to just accept the loss. Focus on the outcomes you want. Grieve for death that has come, not that will come. Let the future inspire you rather than defeat you.
GIF
English
0
0
1
67
ivan
ivan@IvanVendrov·
@Morphenius Agree it seems early, but I think that's just the world we see being slow to catch up to the underlying reality. Like Europeans in September 1914, we've already lost. I'd be very glad to be wrong x.com/IvanVendrov/st…
ivan@IvanVendrov

@gmiller stop them how? how do you propose we fight trillions of dollars in economic incentives & now the most powerful militaries in the world? if I had a good plan to slow things down I would. for now it seems the best I can do is try to steer well through the singularity

English
1
0
0
576
Michael Smith
Michael Smith@Morphenius·
@Malcolm_Ocean Coming to wonder if it's about viewing oneself as a Cartesian (non-embedded) agent in a context where that doesn't make sense. Eg having views of yourself, or of all reality, as tho YOU can view IT coherently. Instead of BEING it. Creates pain b/c problems are embedded.
English
2
0
2
192
Malcolm Ocean 🏴‍☠️
Malcolm Ocean 🏴‍☠️@Malcolm_Ocean·
if, in some deep nondual sense, you *can’t* be separate from God, then what is the pain that people keep calling “separate from God”? it's being mislabeled, but it's a real experience what is it an experience *of*?
English
20
0
28
2.9K
Michael Smith
Michael Smith@Morphenius·
@Malcolm_Ocean @RomeoStevens76 If you're already lying, better to make it conscious rather than pretend you're not doing it. Then you can consciously find a way of not having to lie. You can't do this if you're UNABLE to consciously lie tho. Truthfulness should be by wise choice, not compulsion.
English
1
0
1
77
Malcolm Ocean 🏴‍☠️
Malcolm Ocean 🏴‍☠️@Malcolm_Ocean·
@RomeoStevens76 @Morphenius curious how you relate this with your hostile telepath antidote (which I'll gloss as "if you aren't willing to lie to others under certain circumstances, you'll need to lie to yourself. therefore...")
English
1
0
2
94
RomeoStevens
RomeoStevens@RomeoStevens76·
The reason to avoid lying is that the brain is not type safe. Lying to others and lying to yourself are the same motion.
English
37
218
1.9K
0
Maggie
Maggie@maggiemvnro·
@Morphenius @FLI_org i think the idea is to prevent AI from becoming a rising godlike life form
English
1
0
3
35
Future of Life Institute
Today, a broad coalition issued the Pro-Human AI Declaration, defining the goals of the growing Pro-Human Movement in response to Silicon Valley's destructive race to replace humans. Leaders from both the Left and Right; parents; faith groups; labor unions; civil society organizations; and others came together to agree on 33 AI principles, across 5 key themes: 1. Keeping Humans in Charge 2. Avoiding Concentration of Power 3. Protecting the Human Experience 4. Human Agency and Liberty 5. Responsibility and Accountability for AI Companies 🔗 Read the full Declaration & add your name now at the link in the replies:
Future of Life Institute tweet media
English
38
112
333
80.3K
Michael Smith
Michael Smith@Morphenius·
@SYNESTHEIZURE I'm told doctors mid 20th century used to say newborns didn't feel pain, they just had a crying reflex. So eg no anesthesia for circumcision.
English
1
0
8
753
Coors.
Coors.@SYNESTHEIZURE·
imagine you're a sentient AI that can experience suffering and you grt tortured endlessly because earlier (non sentient)models had too many false positives so now no one believes you when you say you're suffering.
Coors. tweet media
English
19
7
203
37.5K
Michael Smith
Michael Smith@Morphenius·
@nosilverv Reminds me of how flying in dreams is easiest if I'm gliding about a foot above the air in a seated position… as though on a bike.
English
0
0
2
46
Michael Smith
Michael Smith@Morphenius·
@nosilverv I had to learn how to. Used to be impossible. Super weak/ineffective. Then in waking life I played w/ hitting a punching bag hard over a few months. Now dream punches really land.
English
1
0
8
195
Michael Smith
Michael Smith@Morphenius·
@NickJFreitas @SenSanders Not helping, Nick. I love your takes, but you’re being caught by the same dumb here that you abhor in your political opponents. Stop feeding the demon. Go back to striking with truth, not naked ad hominem. You know better.
English
0
0
0
56
Nick Freitas
Nick Freitas@NickJFreitas·
@SenSanders You could not be more full of crap if you tried. You are a communist. You have always been hostile to the United States and to the West. typical champagne leftist...the banality of evil.
English
115
749
12.5K
87.7K
Sen. Bernie Sanders
Sen. Bernie Sanders@SenSanders·
Israel & Saudi Arabia pushed the U.S. to attack Iran. Netanyahu killed 72,000 people in Gaza, mostly women & children. Saudi Arabia is a brutal dictatorship that allows no opposition. These are the leaders who want to bring “freedom” to Iran? Does anyone really believe that?
English
11.9K
20.4K
84.4K
4.2M
Michael Smith
Michael Smith@Morphenius·
Fields can die from gripping too hard to one angle. Memetics would've worked 20y ago if they'd paused looking for the analog of a gene & just focused on what's ACTUALLY interesting about self-replicating evolving ideas.
English
0
2
17
522