I actually think the central thesis is thought provoking, we have shifted far away from locally installed shit to remote data centre access, this was initially driven by cloud-based initiatives and now spiralling upwards by AI. For any researchers, hackers, builders wanting to play with locally installed AI, hardware could become a bottleneck especially as many machines, such as the beloved Macs, are not upgradable
Case reports are the lowest levelnof evidence for a reason. Troublenis when sustematic reviews of a certain conditin is based on historical case reports and noise turns into pseudo-signal
There have been some great cases reports, Broca’s area, but you must scrutinise everything youn read.
First there is no 'one' cancer
There are cancers, subtypes of cancers, subsequent molecular variants and so on..
Therefore, there isn't a need for ONE cure, but a whole arsenal of options to be able to tackle them all. And many subtypes are just too nasty, they are a death sentence unfortunately.
There are endless teams led by principal investigators (PIs) who invest their entire careers reading loads of peer-reviewed scientific articles, formulating hypotheses, applying for research grants, hiring staff, conducting research, publishing their own peer-reviewed papers, presenting at conferences, forming international collaborations with other PIs to expand their output, and generally try their best to push the conversation forward in search of solution, potential molecular / therapeutic targets. There are then clinical trials, treatment protocols etc The fight against 'cancer(s)' is a concerted effort, moves slow due to the super complexity of the matter. There are novel drug discovery pipelines, some leveraging AI such as Isomorphic labs, that promise to accelerate treatment discovery. The next 20 years we might see things getting accelerated. But for beings who have been evolving biologically over a very very long time, millions of years, spending 30+ years for the discovery of a single cure of a certain subtype of cancer is seen as a short time. That might change soon, but the tasks is so ridiculously difficult we don't need silly conspiracy theories to make sense of it.
It is a philosophical Q though, if that logic can flow downstream, then who says WE are the upstream but not somewhere further up? 'Everything is flux'
Open by Andre Agassi. Intriguing to read on the struggles of a tennis legend which were concealed at the time. He provides great detail and insights. It is a great follow up from 'The inner game of tennis', speaking of how athletes, and anyone else really, can find punching themselves down through negative mindset which if overcome can unlock your true potential.
What are peoples' favourite md implementations? Curious as there are different varieties and even more varied opinions. I am building a lightweight project folder managing app supporting markdown and I am between Commonmark and GitHub flavoured markdown and want to gather thoughts.
'.. if a value-aligned, safety-conscious project comes close to building AGI before we do, we commit to stop competing with and start assisting this project. '
Which such project is that, though? And would it accept OpenAI's assistance?
AGI, having access to our world, is precarious as alignment with humans is never guaranteed. Having a buffering medium, aka a simulation environment where AI operates might be a better in-between solution.