TOP 500 AI Game
Could you help me? The easy part of this AI Game is a listing of 500 AI's (well there are many such lists); the hard part is updating premier league top 10 showing how each win-win with each other pairwise or in other combos -and how this chnages what colege students now need to unite across disciplines. There is no point in any 4 year degree graduation if it hasnt included at least a few weeks intelligence cubed updates
I have spent much of my professional life opinion polling and coding (eg back in 1985 my team coded first data on 15000 hours of interviews on what low cost dinsfectant majority of people in india wanted as first nationwide (tv added) brand in that market - data included city slum dwellers and many different dialects). In addition to the interviews and their digital taping in binary, it rfequired about 300 hours of human analysis work (that's before we passed the written report over to the client a partnership between a leading us company and a maharjah) and was limited to testing about 10 possible formulations, price points, - if we had today's AI tools we could have done quite a bit better (but is was as good as my 1985 brain and tools avialable then could do). And overall the client thanked us both at time and when I did a 5 year after review.
so today i have to say ouch- yesterday in washington dc i heard quite the silliest survey ever of what do people feel about AI- it was silly because I dont currently know any 2 people who mean the same thing by ai so all the aggregated numbers were sort of whatever you wanted them to be as 4 panel members droned on for 90 minutes; it bothered me particularly as I had spent much of the weekend at probably America's Main Street (Penn ave) deepest 2 day event of indian and american medical workers sharing inteligence needs with each other
CSPAN India's US Ambassador Vinay Mohan Kwatra
MAY I SUGGEST: AGI MAY BE FAR FAR AWAY
it is quite possible that what ASI does today is play 50 of most intelligent market games ever played plus one sillier one - the sillier one is general chat (before resoining is untegratewd) which I dont think it pays any attention to but is forccd to play by some mdia barons who seem to want to own web3 (whatever that is!)
here's why I sort of trust ai to say its happy hosting 50 inteligence games but isnt anywhere near attaining AGI nor chatting other than a trivial pursuit
back in 1956 Neumann_Einstein-Turing all parted earth within 2 years of each other but listed these final cluses
expect 10**18 mre tech by 2025
design win-win trading gamesolut of every first mile community not just top-down
please dont make ai gurus nations' most hated and wanted of brains the way you mahe energy engineers since einstein's 1905 publication of e=mcsquared ; also if you are going t
***o end wars you will need to end top secrets; and you will also need to be transparent as to when is open source better than deep patenting
start designing computers to improve human brains applying neural networking maths
Instead we hunans coded linearly with binary right up to 1993 when give or take a year 2 things happened jensen huang started up Nvidia to build non linear comouting machines which for 10 years hosted games; yann lecun got neural network algorithms ti replace manual work of checking post codes and cheques signing
TRUMPIST FREEDOM OF VOICE 1993 REPLAY?
Would americans have preffered Yann had not done this and a million clerical jobs had remained in clecking post codes and cheque singatures? - I guess not but its true nobody organised a nationwide vote - in other word intelligence engineers play (with 10**18 more tech) at games beyond ones that public srvants can't regulate- this seems to me to introduce a question on how media covers this which I havent yet seen commercial media address
2002 Beyond )01 Coded Computational Machines
anyways, a decade after jensen and yann had started bridging the generation gap (40 years of not proressing the NET's advice) in 2002 steve jobs Pixar and jensen started coding pixels instyea dof binary - it soon became evident that health ai games depended on pattern recognition I would say at least one third of value creating ai games toiday are pattern maths (see eg biotech;s 250 million protein gameboad of alphafold 3 first trained on oriental game go) - jensen first chatted to demis hassabis about this in 2009 and stanford deep learning lab was begun in the engineering quadranngle; in the same year fei-feil li chatted to jensen- why dont you help test if machines can be coded in all 5 human senses? It was 2012 that li and jensen had provided both computing tools and annual world comeption to prove machines could be trained to host 5 heuman sense gamss; every big investor crioded in- google took main share aquiring hassabis deep mind )pattern games); oddly enough MusK & Li & Jensen thought that mobile autonomus games most needed to wait until 5 human senses were machine supercomputable- this is what was announced GTU 2015 if not earlier; and 10 years later nvidia's annual update climaxed with today nvidia, hassabis goigle deep mind, and disney (todays owner of pixar) annouunce Newton physical AI (essential leap before letting autono,ous humanoids roam your city)
i find it fascinating to look around the workd at 50 most dinstinctive partners of jensen or lecun or hassabis logic as well as fei0fei li;s intennts ; toghether with thise 4 extraordunary enginners (and thousands of alumni's) contributions i like what king charles; japan emperor; modi and bloomberg hope for their peoples emerging brainpower (3 nations including youth biggest prain poulation) and 45 cities learning with bloomberg since 2007. More at www.economistjapan.com/2025 on how jf kennedy, and the 2 royal families of UK and Japan were first to take the NET's advice to heart
No comments:
Post a Comment