probability machine. It’s an old idea that computers got fast enough to make acceptable. This technique is highly dependent on training data, meaning topics, languages, and fields with lots of available data and prior art work well. As do highly-structured and typed domains and languages. Now, mix these machines with layers and layers of good old fashioned code.
technology? Sure. Magic? no. It’s not Artificial General Intelligence (AGI), and the current approach likely won’t get us there It’s also not purely hype It’s not immune from market forces
years Where we’ve been before Personal Computing The Internet A computer in every home. Entire new jobs created to work on computers instead of paper. Killer app: The Spreadsheet A network of computers that could talk to one another. And entirely new media ecosystem and economy. Killer app: HTML and the hyperlink Mobile Always on, and always connected society. Computing power in your pocket, and an entirely new app economy. Killer app: App stores
Money flows into the system. Experiments everywhere. Exuberance. Bust: Money comes due. Stuff gets real. Winner and losers are decided. Absorption: Technology enters the mainstream and gets integrated into society at a background level. A theory: AI is likely in the same cycle
system. But, it’s increasing circular - with the same money flowing from invested party to other invested party Effectiveness of current tools reaching diminishing returns Business models feeling the strain as bills start to come due and investors want returns
dummied
shift_evolve
hayama17
miichan
daitak
syucream
yama3133
hiranabe
yoshimi0227
seike460
knih
jnchito
kazzpapa3
madoxten
nikkihalliwell
ipullrank
livdayseo
baasie
jcasabona
bluesmoon
marktimemedia
raygrieselhuber
tmiket
codingconduct
gurzu
