@elonmusk@elonmusk I gotta say that you have always been a hero of mine. You have accomplished more than anyone in the entrepreneurship world.
But your new alignments are mislead. This kind of shit is simple ridiculous. Please, stop supporting ambushes like that by clowns like Trump...
“O Congresso de TI 2024 não é só sobre tecnologia, é sobre pessoas. Quer entender tendências e crescer na área? Comece agora.”
👉 Inscreva-se para a 11ª edição: https://t.co/unGuy8G3E3
Olha que sacada fod@ pra vender mais na Black Friday... O AppSumo usou uma jogada simples, mas poderosa: ao rolar até 75% da página, aparece um botão que te diz, 'Pressione B para comprar'. Parece básico, mas é genial. Que mudanças você faria no seu site pra otimizar as compras?
O marketing mudou, a tecnologia avançou, mas muita gente ainda está no passado. Amanhã, às 18h45, vou discutir o futuro do marketing com alguns experts. Se cadastra em https://t.co/kd3Qj5rHrb
Muito bacana falar sobre os Motions de Go To Market pra uma turma de empreendedores tão fod@s quanto vocês!
Parabéns @inovenow pela iniciativa, bora crescer!
#gtm#growthmarketing#growthhacking
Destroying Terrible Anti-AI Arguments: Day Four
"My p(doom) ration is X%"
Nate Soares, the president of the Machine Intelligence Research Institute, (an AI research institute in the same way that the People's Democratic Republic of North Korea is a democratic republic) wrote that there's a 95% chance that AI will kill us all and all AI research must be stopped now before it's too late!
Where does this 95% chance number come from?
Nowhere. It's completely made up.
What is the actual probability that AI will kill us all? Trick question! Any answer someone gives is just totally made up based on absolutely nothing more than someone's feelings.
How people come to this number is based on their personal experience, relative intelligence, the things they read, whether they were bullied as a child, their general chemical makeup, whether mommy was nice or mean to them, where they want to school, what they saw, and all the other little things that make up who we are as a person over time.
What it is not based on is any actual predictive model of the future.
People are experience machines and we build our models of reality based on that and we interpret reality through those models. There is nothing to say that model is right. We are a black box and we don't always generalize well. People generally never experience reality directly, but through the filter of their senses and their mind's predictions about what is happening. We see a snapshot of what is around us and fill in a lot of the blanks.
But there is no actual predictive formula, no statistical model, no actual rubric for getting to that number.
It's just people putting their finger to the wind and saying "I think it feels like 20%." They're making up a number based on absolutely nothing more than a feeling and a personal conviction that it's a real threat.
He also wants you to know that there's a "debate within the alignment community re[garding] whether the chance of AI killing literally everyone is more like 20% or 95%" and that he is in "the more like 95% myself" camp. That's just an appeal to the crowd and a false one at that. Actual polls taken, such as the Ipsos Global Views on AI 2023 report, show that 54% of people across the world in 20 major nations from the US to Thailand to India, say "AI has more benefits than draw backs" with old folks skewing towards fear (43% in Boomer) and younger folks skewing much more positive (62%). Either way, what the crowd actually believes is of zero use to us anyway and makes no difference in clear, sound, critical debate because the crowd can be wrong.
Also, the AI-will-kill-us-all contingent is just a small echo chamber faction of the alignment community, which is itself a tiny fraction of AI researchers, which is itself a tiny fraction of people on Earth. And their numbers are no more real than a number you make up about the probability of aliens invading tomorrow or a coronal mass ejection from the sun killing us all.
Saying your p(doom) ratio is also a trap set by the person asking it. If you say anything above zero then the person immediately jumps to "well if there is any chance that the apocalypse can happen then we've got to do everything we can to stop it now!"
Unfortunately, everything good and worth having in life is a risk. You cannot make things "absolutely 100% safe." There is nothing in life that is guaranteed. You can sit inside all day like an Agoraphobic who never leaves the house and you still aren't safe.
Life is a risk. It's a risk every time you get into a car. You have a 1 in 93 chance of dying in a car accident in your lifetime. You still get in a car. It's worth the risk because walking everywhere is great when everything is a few miles away but not so great if you want to go from New Jersey to New York.
Not using or allowing AI is also a risk. Take driving, for a great example. Why do we want AI to drive? Because humans are terrible drivers. We kill 1.35 million people on the road every year and injure 50 million more. If AI cuts that by half or down to a quarter that is 1 million people walking around and playing with their children and living their lives.
The default belief of many in the AI-is-dangerous movement and that in the future it will become so dangerous that it will rise up and murder us or cause a catastrophe. Not just metaphorically. They believe in total extermination of human beings, literally. They use that word a lot, literally, because they really mean total extinction of the human race. In other words they watched Terminator too many times and now believe it's reality. Their basic idea is that intelligent and nice don't always line up. Unless we can guarantee that AI will always do exactly what we want, we shouldn't take any chances! Better safe than sorry!
But again, where does this belief come from? We like to think of our beliefs as sacred but when our beliefs don't accord with reality what good are they? You might believe that gravity doesn't exist but if you go step out of a 100 story window you will see that gravity has an undefeated record. These beliefs on AI's future danger are based on speculation, imaginary concepts and things that simply don't exist. And to be very clear, humans are generally not great at prediction. We are great at imaging all the things that will go away but terrible at predicting the new developments and the good things that will happen.
You cannot predict the Internet and what it will be like as an 18th century farmer. That's because it's built on the back of a chain of inventions like electricity, wires, computers, semiconductors, information theory, the browser, graphic design, and so much more. It's impossible to see all of that coming together and how it will effect life in any meaningful way. The only way to understand it is to get there in time and experience it.
We have no superintelligent AI (in the way people commonly imagine it) and don't know what it will look like *if *and when it does arrive. Actually, we can make the case that we do have "superintelligence" already and it looks nothing like what we expected. GPT-4 knows 100s of languages and can speak well and widely on more domains that any single human. If you've mastered three languages you are in the 1% of learners, and ten languages puts you in the upper 1% of 1% of language learners. The Guinness World Record is held by Ziad Fazah who spoke 58 but probably not as well as GPT 4. Our understanding of what we thought AI would look like and how it is actually developing in the real world don't accord, as if so often the case.
A lot of people spend a lot of their time (when they could be doing useful things instead) speculating (badly) about the nature of superintelligence and what it will be like. This is a bit like the Wright brothers speculating about what a Boeing 787 Dreamliner will be like, as well as what society will be like around it. There is no chance that it will be accurate because by the time the 787 gets here it's built on a chain of other developments, technologies and innovations like new materials, computers, electricity, engines, etc. Not to mention the fact that you don't know what society around that Dreamliner will be like because it is built on the back of millions of other developments and inventions and political changes.
And again, why do we assume that true superintelligence AI (if such a thing is even possible to build) will be a homicidal maniac? Why do we assume it will want to wipe out the world? Why do we assume it will have sentience and goals of its own? Why do we assume that there will be no developments in alignment or transparency or understanding or control? Why do we assume there will be no other frictions or developments in society at large to allow us to adapt with the technology, like we have done with 100% of other technologies before it?
Many of the negative answers to these questions are all taken as a given by anti-AI apologists and these presumptions cannot be taken as a given. They exist nowhere outside the pages of sci-fi and sci-fi is a terrible predictor of actual reality since stories are about one thing and one thing only: conflict. That's why all the "AI" we've had in stories are robot buddies or the Big Evil Bad in a metal skin. They are just characters put into a metal sheath. But real AI is developing in ways that simply do not mirror sci-fi up until this point. Sci-fi is not reflective of how technology develops in reality. They're stories, just stories.
On the flip side, it sure seems like more intelligence, more widely spread, is the best way to make this world safer and more abundant for as many people as possible.
When you think about it, what doesn't benefit from more intelligence?
Is anyone sitting around thinking I wish my supply chain was dumber? I wish cancer was harder to defeat. I wish drugs were harder to discover and new materials were harder to create. I wish it were harder to learn a language.
Nobody.
More intelligence is better nearly across the board. We want an intelligence revolution and intelligence in abundance.
If we let it proliferate and we don't kill it in the crib based on regulatory capture disguised as safety or the paranoid worries of people who believe p(doom) is 95% because they read too much sci-fi as a kid, intelligence will weave its way into every single aspect of our lives, making our economy strong and faster, our lives longer and our systems more resilient and adaptable. Where does that prediction come from? Easy. It comes from looking at the entire history of technology and societal development up until this point.
As Matt Ridley wrote in the Rational Optimist: "Since 1800, the population of the world has multiplied six times, yet average life expectancy has more than doubled and real income has risen more than nine times. Taking a shorter perspective, in 2005, compared with 1955, the average human being on Planet Earth earned nearly three times as much money (corrected for inflation), ate one-third more calories of food, buried one-third as many of her children and could expect to live one-third longer. She was less likely to die as a result of war, murder, childbirth, accidents, tornadoes, flooding, famine, whooping cough, tuberculosis, malaria, diphtheria, typhus, typhoid, measles, smallpox, scurvy or polio. She was less likely, at any given age, to get cancer, heart disease or stroke. She was more likely to be literate and to have finished school. She was more likely to own a telephone, a flush toilet, a refrigerator and a bicycle. All this during a half-century when the world population has more than doubled, so that far from being rationed by population pressure, the goods and services available to the people of the world have expanded. It is, by any standard, an astonishing human achievement."
Averages can hide the outliers but if you really look at it, it's nearly impossible to find any place on Earth that is not better off now than it was in 1955 or earlier, because of technology and our collective advances. If you want to have solid grounds for predictions then look to the past instead of predicting sudden and total decline.
In the 1960s the book The Population Bomb predicted that we'd need to let 2 billion people starve because there just weren't enough resources to go around. Instead we got the Green Revolution and can sustain more folks than ever.
Historian Michail Moatsos estimates that in 1820, just 200 years ago, almost 80% of the world lived in extreme poverty. That means people couldn't afford even the tiniest place to live or food that didn't leave them horribly malnourished. It means living on less than $1.90 a day in 2011 prices and $2.15 in 2017 prices. Actually you don't even need to go back that far. In the 1950s, half the world still lived in extreme poverty.
Today that number is 10%.
As Thomas Babington Macaulay wrote, "On what principle is it, that when we see nothing but improvement behind us, we are to expect nothing but deterioration before us?"
I fully support everyone's right to believe whatever they want to believe and to make up any p(doom) number they want by sticking their finger to the wind. But I don't support their right to make stupid and paranoid public policies for the rest of us.
If you believe in an AI apocalypse that's fine but you don't get to make the rules for the rest of us, just as Ted Kaczynski doesn't get to shut down society and send us all back to our imaginary agrarian roots just because he thinks it's a grand idea.