Ben Goertzel- AI Against Aging
Modern science leaves little doubt that death due to aging is a solvable problem. The human body is a complex machine, and it is modifiable and reparable like any other machine. But though the roadmap to radical human healthspan extension is clearer than ever before, the magnitude of the challenges involved is also apparent. And the worst bottleneck we face in surmounting these challenges is: The effectiveness of the brains of human researchers. The human brain simply was not evolved for the integrative analysis of a massive number of complexly-interrelated, high-dimensional biological datasets. In the short term, the most feasible path to working around this problem is to supplement human biological scientists with increasingly advanced AI software, gradually moving toward the goal of an AGI (Artificial General Intelligence) bioscientist. In my talk I will spell out what this would concretely mean for biomedicine, and how we might get there from here via a plausible path of AI technology development. And I will also discuss some things that advanced AI technology is doing for bioscience right now, paving the way for these more radical future possibilities. The Biomind AI tools my colleagues and I have developed have helped discover the genetic basis of Chronic Fatigue Syndrome, and to create novel diagnostics for Parkinson’s Disease based on identifying subtle patterns of damage in mitochondrial DNA. In a current collaboration with Genescient Corp., these same AI tools are being used to analyze the genomes of flies that were experimentally evolved to live 5x as long as ordinary flies, in order to understand why the flies live so long and draw lessonstherefrom regarding which drugs and supplements may best enhance human healthspan. Today’s AI systems are already comprehending the biological world in ways far beyond the human brain’s capability, but they illustrate only a tiny fraction of what’s to come.
Goertzel has an article up at Hplusmagazine
Cost to develop drugs is about $1 billion.
Baked into the cost of drugs is R&D for new drugs.
Genescient methuselah flies -actually super flies. Stronger hearts, better atheletes and are better, more active and tougher in many ways
The differences between superflies and ordinary flies involve 1200+ genes and multiple coupled pathways and networks.
How to gain greater insights using AI tools beyond regular statistical analysis.
Biomind LLC advanced AI for postgenomic bioinformatics
narrow AI and machine learning for data mining.
CAn use AI text mining can integrate knowledge presented in multiple online publications.
A few dozen key genes are the most important to longevity. these are important and get control that requires understanding of the systems. Then understand the longevity hubs and hook it up to drugs and nutrients.
the free ingredients that were identified as key from the AI are familiar and suggests the analysis is not BS.
selenium, vitamin E, sodium selenite, quercetin, genstein, zinc, bioflavins, folic acid, calcitriol, estradoil, resveratrol
Sequencing of the methuselah flies is underway. Results in Fall of 2010.
Current AI tech gives out about 1/10 to 1/100 of what is there.
We need a full fledged generally intelligent artificial biologist
Future of biomed science and other science
animal/child like AGI + narrow AI to get to artificial biologist