Ralph Cerchione
@dry-observer.bsky.social
"Knowledge is a single point, but the ignorant have multiplied it."
Here's my video on OpenAI's o1/Strawberry Model, and the runaway technological progress it could foster. In the right hands.
https://www.youtube.com/watch?v=kt4bTzPOLo8
Here's my video on OpenAI's o1/Strawberry Model, and the runaway technological progress it could foster. In the right hands.
https://www.youtube.com/watch?v=kt4bTzPOLo8
o1 spent all of 10 seconds considering the question, and came up with a potential 15-year timeline to Singularity.
Given the pace we're seeing lately - *after* it was trained - I think the time frame is considerably shorter.
Given the pace we're seeing lately - *after* it was trained - I think the time frame is considerably shorter.
January 20, 2025 at 1:24 AM
o1 spent all of 10 seconds considering the question, and came up with a potential 15-year timeline to Singularity.
Given the pace we're seeing lately - *after* it was trained - I think the time frame is considerably shorter.
Given the pace we're seeing lately - *after* it was trained - I think the time frame is considerably shorter.
The incoming Administration may want to use fossil fuels, but that would take an act of Congress. President Biden is using existing law which allows this kind of work on Federal lands so long as everything runs on carbon-neutral energy.
This is an incredible gift.
We shall see what comes of it.
This is an incredible gift.
We shall see what comes of it.
January 17, 2025 at 11:55 PM
The incoming Administration may want to use fossil fuels, but that would take an act of Congress. President Biden is using existing law which allows this kind of work on Federal lands so long as everything runs on carbon-neutral energy.
This is an incredible gift.
We shall see what comes of it.
This is an incredible gift.
We shall see what comes of it.
There is a revolution underway, and we didn’t need AGI or even fully autonomous AI to spark it.
The next question, of course, is how to adapt to revolutionary change.
The next question, of course, is how to adapt to revolutionary change.
November 30, 2024 at 5:44 AM
There is a revolution underway, and we didn’t need AGI or even fully autonomous AI to spark it.
The next question, of course, is how to adapt to revolutionary change.
The next question, of course, is how to adapt to revolutionary change.
But like handing off vast swathes of scientific research to automation, letting your brightest minds focus on the most-critical work instead of grinding out every step only empowers you further and speeds everything forward.
November 30, 2024 at 5:42 AM
But like handing off vast swathes of scientific research to automation, letting your brightest minds focus on the most-critical work instead of grinding out every step only empowers you further and speeds everything forward.
Fully autonomous AI isn’t necessary to do this work - remember we’re on a planet with over 8 billion people, billions of whom already have smartphones with AI upgrades, and millions of whom have the skills you need.
November 30, 2024 at 5:41 AM
Fully autonomous AI isn’t necessary to do this work - remember we’re on a planet with over 8 billion people, billions of whom already have smartphones with AI upgrades, and millions of whom have the skills you need.
When LLMs struggled with math, we gave them a calculator. When Google DeepMind’s LLM was too unreliable to solve geometry on its own, they partnered it with a rule-bound deduction engine.
November 30, 2024 at 5:40 AM
When LLMs struggled with math, we gave them a calculator. When Google DeepMind’s LLM was too unreliable to solve geometry on its own, they partnered it with a rule-bound deduction engine.
And then hovering over all of this are AGI and its precursors.
Remember, even before we reach full AGI, if you have an intelligence which can oversee all these constituent parts and leverage all these breakthroughs, you have a system which can accelerate their deployment.
Remember, even before we reach full AGI, if you have an intelligence which can oversee all these constituent parts and leverage all these breakthroughs, you have a system which can accelerate their deployment.
November 30, 2024 at 5:39 AM
And then hovering over all of this are AGI and its precursors.
Remember, even before we reach full AGI, if you have an intelligence which can oversee all these constituent parts and leverage all these breakthroughs, you have a system which can accelerate their deployment.
Remember, even before we reach full AGI, if you have an intelligence which can oversee all these constituent parts and leverage all these breakthroughs, you have a system which can accelerate their deployment.
And this is just one technology, CRISPR, applied to one technique, gene therapy, in one field, medicine.
AI will help us understand our genes better, and how epigenetics further impacts them, but so will mass, parallel testing using tools like microfluidics and 3D printing “labs on a chip.”
AI will help us understand our genes better, and how epigenetics further impacts them, but so will mass, parallel testing using tools like microfluidics and 3D printing “labs on a chip.”
November 30, 2024 at 5:28 AM
And this is just one technology, CRISPR, applied to one technique, gene therapy, in one field, medicine.
AI will help us understand our genes better, and how epigenetics further impacts them, but so will mass, parallel testing using tools like microfluidics and 3D printing “labs on a chip.”
AI will help us understand our genes better, and how epigenetics further impacts them, but so will mass, parallel testing using tools like microfluidics and 3D printing “labs on a chip.”
The steady elimination of genetic flaws will eventually remove every inborn weakness weighing down each of us, while improvements in both the technology and our understanding of genes will lead to augmentation - selecting genetic advantages to our health, intelligence and other traits.
November 30, 2024 at 5:26 AM
The steady elimination of genetic flaws will eventually remove every inborn weakness weighing down each of us, while improvements in both the technology and our understanding of genes will lead to augmentation - selecting genetic advantages to our health, intelligence and other traits.
This should drive a self-funding boom in gene therapies, as the savings roll over from each major disease cured will enrich the researchers and aid the overall economy.
November 30, 2024 at 5:22 AM
This should drive a self-funding boom in gene therapies, as the savings roll over from each major disease cured will enrich the researchers and aid the overall economy.
However, we need to realize when AGI arrives, as momentous as it is, it may be more of a milestone and a catalyst on the way to superintelligence than the one essential key.
And it may already share the planet with indisputable superintelligence once it gets here.
And it may already share the planet with indisputable superintelligence once it gets here.
November 30, 2024 at 5:16 AM
However, we need to realize when AGI arrives, as momentous as it is, it may be more of a milestone and a catalyst on the way to superintelligence than the one essential key.
And it may already share the planet with indisputable superintelligence once it gets here.
And it may already share the planet with indisputable superintelligence once it gets here.
And regardless, we definitely have organizational and speed superintelligence already in our world, and making their presence felt.
We haven’t reached AGI yet - apparently.
But we will.
We haven’t reached AGI yet - apparently.
But we will.
November 30, 2024 at 5:14 AM
And regardless, we definitely have organizational and speed superintelligence already in our world, and making their presence felt.
We haven’t reached AGI yet - apparently.
But we will.
We haven’t reached AGI yet - apparently.
But we will.
And that’s leaving aside the most basic, such as adequate sleep, nutrition and exercise, not to mention being well read, insightful, and motivated.
Whether or not any lone biological human has hit superhuman capabilities as yet is unknown, but they probably wouldn’t announce it if they did.
Whether or not any lone biological human has hit superhuman capabilities as yet is unknown, but they probably wouldn’t announce it if they did.
November 30, 2024 at 5:12 AM
And that’s leaving aside the most basic, such as adequate sleep, nutrition and exercise, not to mention being well read, insightful, and motivated.
Whether or not any lone biological human has hit superhuman capabilities as yet is unknown, but they probably wouldn’t announce it if they did.
Whether or not any lone biological human has hit superhuman capabilities as yet is unknown, but they probably wouldn’t announce it if they did.
Or the benefits of using sensory deprivation tanks, cranial electro-stimulation (CES), transcranial direct current stimulation (tDCS), biofeedback and similar technologies, which tend to work synergistically with mental enhancements and biotech augmentations like nootropics.
November 30, 2024 at 5:10 AM
Or the benefits of using sensory deprivation tanks, cranial electro-stimulation (CES), transcranial direct current stimulation (tDCS), biofeedback and similar technologies, which tend to work synergistically with mental enhancements and biotech augmentations like nootropics.
Or an AI EXOcortex providing constant augmented reality assessments of your environment while instantly tackling any task or challenge which presents itself without being asked.
November 30, 2024 at 5:03 AM
Or an AI EXOcortex providing constant augmented reality assessments of your environment while instantly tackling any task or challenge which presents itself without being asked.
Essentially anything which improves a person’s physical or mental capabilities counts.
Consider biotech interventions like nootropics (smart drugs), genetic engineering or cybernetics.
Or AI assistance such as chatbots handling coding or data analytics for you.
Consider biotech interventions like nootropics (smart drugs), genetic engineering or cybernetics.
Or AI assistance such as chatbots handling coding or data analytics for you.
November 30, 2024 at 5:01 AM
Essentially anything which improves a person’s physical or mental capabilities counts.
Consider biotech interventions like nootropics (smart drugs), genetic engineering or cybernetics.
Or AI assistance such as chatbots handling coding or data analytics for you.
Consider biotech interventions like nootropics (smart drugs), genetic engineering or cybernetics.
Or AI assistance such as chatbots handling coding or data analytics for you.
If you enhanced someone already operating at a level most exceptional people would already perceive as superhuman, what would improving their capabilities by roughly 10% do?
What about 50%?
What about 50%?
November 30, 2024 at 4:57 AM
If you enhanced someone already operating at a level most exceptional people would already perceive as superhuman, what would improving their capabilities by roughly 10% do?
What about 50%?
What about 50%?
While it’s an open question as to whether we have geniuses of this caliber alive today, much less how many, we clearly do have many people of either genius or near-genius intelligence.
November 30, 2024 at 4:56 AM
While it’s an open question as to whether we have geniuses of this caliber alive today, much less how many, we clearly do have many people of either genius or near-genius intelligence.
So in effect, with minimal schooling and one good book, he didn’t so much as outperform a century’s worth of mathematicians, as outperform them by more than fifty-times over.
What is the point?
These are merely human geniuses and, to the best of our knowledge, not enhanced in any way.
What is the point?
These are merely human geniuses and, to the best of our knowledge, not enhanced in any way.
November 30, 2024 at 4:54 AM
So in effect, with minimal schooling and one good book, he didn’t so much as outperform a century’s worth of mathematicians, as outperform them by more than fifty-times over.
What is the point?
These are merely human geniuses and, to the best of our knowledge, not enhanced in any way.
What is the point?
These are merely human geniuses and, to the best of our knowledge, not enhanced in any way.
In two years, he rederived a century’s worth of mathematics while also developing many theorems original to him alone. One of his many theorems became the foundation of superstring theory in physics, for example.
November 30, 2024 at 4:52 AM
In two years, he rederived a century’s worth of mathematics while also developing many theorems original to him alone. One of his many theorems became the foundation of superstring theory in physics, for example.
Srinivasa Ramanujan, on the other hand, was a young man who could only attend early schooling before he had to go to work to help support his family. He came into possession of an excellent math textbook which was about a century out of date.
November 30, 2024 at 4:51 AM
Srinivasa Ramanujan, on the other hand, was a young man who could only attend early schooling before he had to go to work to help support his family. He came into possession of an excellent math textbook which was about a century out of date.
Leonardo daVinci, the classic “Renaissance man,” was a master and innovator in many fields - anatomy, invention, engineering, geology, painting, sculpture - with many of his accomplishments being centuries ahead of his contemporaries.
November 30, 2024 at 4:48 AM
Leonardo daVinci, the classic “Renaissance man,” was a master and innovator in many fields - anatomy, invention, engineering, geology, painting, sculpture - with many of his accomplishments being centuries ahead of his contemporaries.
Speed superintelligence.
It’s already here, it's just not widely distributed.
And finally, quality superintelligence sounds challenging, but we forget how close we are already.
It’s already here, it's just not widely distributed.
And finally, quality superintelligence sounds challenging, but we forget how close we are already.
November 30, 2024 at 4:47 AM
Speed superintelligence.
It’s already here, it's just not widely distributed.
And finally, quality superintelligence sounds challenging, but we forget how close we are already.
It’s already here, it's just not widely distributed.
And finally, quality superintelligence sounds challenging, but we forget how close we are already.
What Google DeepMind has done is demonstrate that every critical area of study which can be automated by AI ultimately will be.
Are you looking for promising metamaterials? Better designs for aerospace, be they wings, engines or fuselages?
Are you looking for promising metamaterials? Better designs for aerospace, be they wings, engines or fuselages?
November 30, 2024 at 4:41 AM
What Google DeepMind has done is demonstrate that every critical area of study which can be automated by AI ultimately will be.
Are you looking for promising metamaterials? Better designs for aerospace, be they wings, engines or fuselages?
Are you looking for promising metamaterials? Better designs for aerospace, be they wings, engines or fuselages?
And that is only the beginning.
Those five projects are all targeting key areas which can dramatically advance technology - AlphaFold in biotech, and the other four, ultimately, in essentially all of sci/tech.
But again, that’s just a step.
Those five projects are all targeting key areas which can dramatically advance technology - AlphaFold in biotech, and the other four, ultimately, in essentially all of sci/tech.
But again, that’s just a step.
November 30, 2024 at 4:36 AM
And that is only the beginning.
Those five projects are all targeting key areas which can dramatically advance technology - AlphaFold in biotech, and the other four, ultimately, in essentially all of sci/tech.
But again, that’s just a step.
Those five projects are all targeting key areas which can dramatically advance technology - AlphaFold in biotech, and the other four, ultimately, in essentially all of sci/tech.
But again, that’s just a step.