YouTube videos: Text to Image, 168 hour work week, Synchrotron Latin and South America

Two Minute Papers: NVIDIA’s New AI: Text To Image Supercharged! at https://www.youtube.com/watch?v=k4k5RTNX-Js

Your examples are all misleading, when you show only a portion of the input images. Your two minutes went to 6:36. Not to be too cynical, but your examples are a bit “click baity”. In some jobs, this might be useful. But training for the tens of thousands of occupations, I do not think NVIDIA is quite up for the task yet. Things are improving, but they will go faster with more practical challenges and real tasks.

If you add $Thanks, more people can support your good efforts.

Richard Collins, The Internet Foundation


Logically Answered: The 40-Hour Week Is About To Change Forever at https://www.youtube.com/watch?v=BzOmgp1_uwg

Your description is wrong. The standard workweek was 48 hours and Ford reduced that to 40. I work 120 hours a week. I could reduce that to 40 if I had reliable AIs. But then I would go back to 120 because there are 100 times more things that need doing than 8 Billion humans can do. Most organizations responses to Covid and “open global work” were horrible. Many organizations are successful and hire lots of humans. Then they become more efficient as they learn their market and job. But they make more money and hire more humans. Efficiency per person drops and they stagnate. The originators and innovators leave or retire in place. Unlimited growth only comes when the information is stored in open systems, not in a few human heads. And when groups look globally for the many unfinished tasks that are left unfinished. I hope all the human programmers working by the hour are replaced by certified AIs who work 24/7 (168 hours per week) and constant speed and updated with global best methods, always improve. Then those human programmers doing the minimum possible can all retire. Will the AIs take care of them? I doubt it. Richard Collins, The Internet Foundation


Spend a few years, find all the people and groups who are working where this kind of data will be helpful. Facilitate them getting it from existing systems, rather than building one and hoping “they will come”. You have no established networks. You are hoping it will help. But when you actually talk to every person in Latin America who is interested in Science Technology Engineering Mathematics Computing Finance Governance Internet (STEMCFGI) you will find they have ideas already for what they need. Selling them some outsiders technology only benefits the organizers and insiders. Invest longer term and train first. Trickle down does not work, it only enriches the suppliers and existing networks. I am sure you all want to make things better for more people, but you are using bad methods that end up absorbing resources not generating them.

YouTube could support sharing of sheets and tools inline, but I had to copy and paste.

“synchrotron” has 11.2 Million entry points globally

Country TLD Entries
Mexico .mx 10,700
Argentina .ar 5,860
Brazil .br 5,820
Venezuela .ve 3,860
Colombia .co 3,460
Peru .pe 3,370
Cuba .cu 3,260
Bolivia .bo 2,370
Puerto Rico .pr 2,370
Ecuador .ec 1,970
Chile .cl 1,860
Paraguay .py 1,470
Guatemala .gt 1,270
Costa Rica .cr 1,170
El Salvador .sv 1,070
Panama .pa 1,070
Honduras .hn 1,060
Dominican Republic .do 780
Nicaragua .ni 780
Uruguay .uy 780

There are 24,500 results for “synchrotron” (“Latin America” OR “south America”) as of August 2, 2023, 1:05 PDT
It is possible to find all the individuals involved, document their current efforts, help them work together, ask their needs, do not assume.  I did not search all the languages and related terms. Do that and you will at least know who is out there, what they are doing, what they are trying to do, and what else they could do.  The current global synchrotron sites are BAD at sharing data and methods with everyone (5 billion), not just insiders (a few ten thousands). There are many alternate technologies now that are lower cost, less “all your eggs in one basket”, and that encourage local innovation and global sharing. You have not done “due diligence”. There are about 665 Million in Latin and South America and at least 1 in a thousand will try for research STEMCFGI careers.

Brazil (6,320 results)
Mexico (4,270 results)
Argentina (3,860 results)
Chile (2,860 results)
Colombia (2,770 results)
Peru (1,970 results)
Venezuela (1,860 results)
Cuba (1,770 results)
Ecuador (1,760 results)
Bolivia (1,670 results)

Richard Collins, The Internet Foundation

Comment2: Free electron lasers, nanoscale light sources, high harmonic gain extreme ultraviolet and soft x-ray, TeraHertz methods, nano MRI microscopy and spectroscopy, many kinds of accelerators, plasma technologies, high magnetic field technologies, acoustic levitation, 3D laser ionization, and thousands more. Identify all the related technologies, and help built those networks for Latin and South America. Make it all global, so it NOT “Only for Latin and South American’s and exclude everyone else.” Get ALL the groups to share open lossless data, open algorithms and global accessible collaborative sites. Richard Collins, The Internet Foundation


 

Richard K Collins

About: Richard K Collins

The Internet Foundation Internet policies, global issues, global open lossless data, global open collaboration


Leave a Reply

Your email address will not be published. Required fields are marked *