3 ms·
> We see diminishing returns everywhere, and if this is not stopped and we don't manage to drastically increase the efficacy of scientific research we'll effect
by chr1 7y ago
> We see diminishing returns everywhere, and if this is not stopped and we don't manage to drastically increase the efficacy of scientific research we'll effectively flatline at some point.
Most likely there is no way to increase efficacy, because the problems we try to solve become harder and harder as we go. This is why the predictions of singularity are unrealistic, the self improving AI, and constantly increasing number of human researchers will go into merely not flatlining instead of solving everything at once.
- roca 7y agoIndeed! It's especially relevant because being smarter doesn't obviate the cost of doing experiments. The superintelligence still has to build hardware and wait around for results.
- pas 7y agoIt's possible with better models (that make problems look simpler). This part is probably an education issue. Making "smarter" humans (either via sperm/egg/embryo selection, and/or via brain-computer interfaces) would also help a lot. And it doesn't look like we're anywhere close to "peak brain", after all we do a lot with just 20W and ~1.5 kg of wetware. Whatever improvements our potential descendants make will surely look like the big-s-singularity to us.