@emollick Be good to do set of predictions of firsts on a timeline such as "first Ai avatar successfully posing as a human remote worker", first fake photo on the cover of the New York times- this part of Ai isn't going to be good for society
@deedydas I watched a Peter Thiel lecture to Stanford grads where he basically said 'do whatever you can do to avoid competition'. Presume if you can't avoid it then second line is try and starve off the competition, if you have deep pockets offer services as free etc
@AukeHoekstra Do you need to estabish a training threshold? The training miles in US cities are very high and the US grid system is perhaps more easily navigated than non-grid systems?
@krishashok@prakashasj I was turned off Maths at school partly due to my own stumble in trying to contextualise/ ask 'why' about everything. Much better to accept most concepts in Maths as tools and then start practicing (for me at least) -then insight naturally develops later on.
@emollick Most technology acts as a time machine (faster transport, quicker recovery from disease). Ai has arrived after digital (and related easy access to information)- it's not as obvious outside of coding what the application is but the term co-pilot is a very good articulation
@russellcurtis Would it ever be plausible to have a hybrid model with the spanish systems of build roads and amenities to plots but setup for self builds? Seems there's a decent amount of pent up demand for self build opportunities but plots are very limited...
@skdh I think your coming from a scientific perspective whereas others are coming from a socio-political perspective in that governments could have deployed a precautionary principle similar to how the US started to regulate CFC's ahead of definitive evidence.
@svpino Some of the most acutely stressful incidents in my working life have been related to Windows or MS Office breaking or malfunctioning close to a deadline or important presentation
@gmiller We're really boxing with shadows as we don't know what ceilings Ai might hit which will slow things down enough for stable economic equilibriums to be established. Anything regulated will be slower to adopt and general robot capability will lag digital intelligence?
@fchollet .. but, especially in the company space, there's an awful lot of people essentially producing the same code for the same sort of output but everyone is silo'd in their own company.
@emollick I've yet to understand why this is an issue except for those wanting to make an in-penetrable moat i.e. the high compute $ needed to train such massive data acting as brake on competition