1979 & Proyouth GAMES to Linkin from 1951: Ed's & A!20s most curious moments as V. Neumann's & The Economist's diarists include 1982...LLM2022STORY why we co-brand with AIgoodmedia.com.When The Economist sent dad Norman Macrae to pre-train with Von Neumann 1951 Princeton, they agreed The Economist should start up leadership Entrepreneurial Revolution surveys; what goods will humans unite wherever they first linkedin to 100 times more tech per decade? Johnny added a final twist in notes for his biography. "Unfortunately Economics is Not Mathematical. One day only AI maths can save our species

Breaking: help prep AI rehearsal Fringe UNGA Sept 2023 NY- chris.macrae@yahoo.co.uk
July Guterres choosing top20 AIHLAB.. bard says Hassabis will chair this '''''with UN tech envoy ..members include Stanford's Fei-Fei Li , Allen's Etzioni, Sinovation's Kai Fu Lee,... Gemini,,Uni2 :FFL*JOBS*DH more G : 1 2 3 4 5
Guterres*JYK*JFK
..worldclassllm & Royal Family's 150 year survey: can weekly newspaper help multiply trust around worldwide human development?
0: Around WorldMaths #1 FFL in 80.. 79

Game AI : Architect Intelligence:: EconomistDiary invites you to co-create this game & apply bard.solar ; personalise your pack of 52 players cards. Whose intelligence over last 75 years most connects human advancement at every gps concerning you and yours on planet?
we offer 3 types of tours sampling rockstars on intelligence for good and welcome guest tours :Alpha Chronological began 1951 through 4 decades at The Economist; Gamma: back from future of 2020s began 1984; Beta intergeneration connectors are more recent quests; try  AI game out; we'd love to hear whose action networks inspires You and who chris.macrae@yahoo.co.uk
Alpha1 JFKennedy Neumann-Einstein-Turing Crowther; Youth visions for 1960s launched by Kennedy as great as any known to us- eg space race; peace corps, Atlantic-Pacific win-win trade; Kennedy had studied quite traditional economic gurus at Harvard (eg ); served in US Navy Pacific theatre word war 2; he discovered The Economist stories of exciting economic possibilities; these had emerged from editor Geoffrey Crowther ; his 20+ years of editing included 1943 centenary autobiography of Economist- had been a mistake to vision a newspaper helping 20 something Queen Victoria in 1843 transform to commonwealth trading from slavemaking empire; Crowther thought good news media was worth another go; he sent a rookie journalised who had survived being teen navigator allied bomber command Burma to pretrain with Neumann at Princeton year of 1951 as well as interview NY-UN year 6; Neumann explained after spending their lives mainly on the science allies needed to beat Hitler: Neumann-Einstein-Turing wanted a good legacy - digitalisation -see eg Neumann's last lecture notes delivered Yale "Computer and the Brain". There were 4 inter-generational crises the NET foresaw; sorting out energy; designing win-win economics; sorting out worldwide cooperations; everything else UN and multilaterals were being asked to resolve. Neumann trained Economist journalist in the leadership survey : "What goods will humans unite wherever they have early access to 100 times more tech per decade?"
(breakingJy10) Gamma1 Hassabis , Fei-Fei Li,, Guterres, Oren Etzioni, JYKim, Ng, Yang, Chang, Chang- There are lots of alternative Gammas but we start with 2 engineers who transformed AI from 2010 when they furst met at Stanford and discussed FFL's NSF funding of imagenet since 2006; 2 public health servants who in 2016 weren't happy with just talking 17 new UN goals and have been asking AI genii to help digital roadmap UN2 since 2016 and a Taiwanese American in Silicon Valley, a Chinese American In Taiwan and Samsung's Korean who partnered Taiwan's chip making genii; these stories have lots of personal courage as well as brilliance; any reporting errors are mine alone chris.macrae@yahoo.co.uk My family has made 100 trips to Asia from the west but still have no fluency in oriental languages so I am biassed : i believe NOW! that LLMs can connect the best cooperation intelligences ever and urgently map life critical knowhow through every global villahge
Beta 1 celebrates massive web and inter-generational  gifts of Steve Jobs Fazle Abed Mr Sudo JYKim and Mr Grant; you will probably know Jobs started 2 digital networking revolutions with 1984s Mackintosh Personal Computer and apple and 2007's iphone; at bottom of pyramid, you may not know Asia-66-percent-of%20Intelligence-for-good-part-1.docx   fazle abed linked up to 1 billion tropical Asian real housewives & entrepreneurs towards  empowering the end of poverty; and Steve hosted silicon valleys 65th birthday party for abed in 2001; they brainstormed transformative education which the pc hadn't delivered ..but could the mobile era be visioned to do so?; Mr Sudo had partnered Abed and Bangladesh villagers in "leapfrog" mobile experiments starting 1995. By 2001, as Jobs was introducing Abed to eg Stanford friends, Kim had discovered Abed's women were networking the most effective solution to rural Tuberculosis; he introduced Gates and Soros to Abed as all 4 wanted 2000s Global Fund to end TB & HIV & Malaria; at the same time Guterres had moved from Portuguese prime minister to red cross and then UN servant leader of refugees; meanwhile back in 1980 it was UNICEF's James Grant who had discovered Fazle Abed women's oral rehydration network which was saving lives of 1 in 3 infants who previously died of diarrhea in the tropics' humid villages ; Grant became worldwide marketer of how parents could mix water sugar and salts as the life saving cure of ORD; naturally James Grant College of Global Public Health has become cornerstone of all the new university cooperations Abed and Jobs started brainstorming in 2001
here we discuss why 73 years as biographers of V Neumann's future visions suggests its critical to map intelligences who got us to 2020s and today's giant co-leapers Gamma-tours; this also opens door to which intelligences at national or other place levels contribute what? - see our 60+ years of intelligences, and eg discussion of why to end extreme poverty we need one open global university of poverty
Beta2 : NB how different scope of 2020s AI is from cross-selection of web2,1 engineers of last quarter century- NB valuetrue purpose of gamifying Architect Intel : borderless engineering can help humans vision 2020's co-creation of web3 and millennials development beyond extinction. Kai Fu Lee, Ng, Melinda Gates, Koike, Lela Ibrahim, Jobs, Satoshi ,Houlin Zhao, Allen, Musk, Brin ,Page , Bezos, Ma, Zhengfei, Torvaulds, Berners Lee, Masa Son, It would be a pity if short-term nationalism stopped us 8 billion humans learning from these tireless innovative beings. Do sub in your regional counterpart. Also note what no conventional strategist saw as Intelligence possible before 2017. To clarify: start with kai fu lee- his best seller on AI in 2017 doesn't explain the ai thats changing every possibiliity of the 2020s but does it good job of AI up to 2017. He also has unique view because he was sent by google to explore china, falling ill at same time as google exiting china, writing up ai that inspired reinventing himself as both venture capitalist in the midst of asia's most extraordinary student suburb (Zhong...) and as curious observer. I see Ng, Ms Gates. Koike, Ibrahim -as civil education heroines/heroes - who are yours ? Satoshi, Zhao, Allen, Musk - gamechangers taking on conflicts that journey us all through tipping points. One day the world may decide it was a blessing that a corporate like google and a revolutionary uni like Stanford co-habited the same 100 square miles- is there any other comparable 100 square miles of brainworkers for humanity. (I love Hong Kong but thats its own story). The other 5 kept digital movements alive -they merit being valued as engineering heroes before you decide how to translate systemic components to your regions' -and mother earth's - urgent needs.

Saturday, July 1, 2023

ai's own top 50 models and advisory boards

help us update top 50 - no particular order; we partly promoted why teens & teachers gain from seeing these models exist -see 40 years of work hypothesising transformative education essential to millennials preventing extinction (73 years overall since Economist (Entrepreneurial Revolution Systems) jourmalists trained with Von Neumann)

models

·AlphaFold 2 Number 1 Science breakthrouh from Hassabis DepMin/google brain a protein folding model www.nature.com

regarding Large Language Models- caused a stir among those waitimg for metaverse (augmented reality goggles) to start web3 designs for humanit- its now lcear web3 has begun with Gateway LL models (which of thousands will be the main gateway is to be discovered - see also llm-web3-human-race.docx and 100 AI curiosity chats at bard.solar ; we believe places need best possible ai advisory boards - see benchmarks right hand col


· ChatGPT-3: one of 3 leading LLM models out of USA large language model that can generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way. This could be used to help teens learn new languages, write better essays, and create more engaging content.  Opens in a new windowwww.analyticssteps.com

· DALL-E 2:  is a text-to-image diffusion model that can create realistic images from text descriptions. This could be used to help teens with their art projects, or to help them visualize new ideas .Opens in a new windowen.wikipedia.org 


· Bard: Bard is one of 3 leading large language model out of US that is still under development, to perform many kinds of tasks, including

oFollowing instructions and completing requests thoughtfully.

oAnswering your questions in a comprehensive and informative way, even if they are open ended, challenging, or strange.

oGenerating different creative text formats, like poems, code, scripts, musical pieces, email, letters, etc. www.engadget.com

· DeepMind ControlSuite: is a suite of AI models that can be used to control robots and other physical systems. This could be used to help teens learn about robotics, or to help them develop new ways to interact with the physical world. neptune.ai

· OpenAI Five:  is an AI team that was trained to play the game Dota 2. It is the first AI team to defeat a professional Dota 2 team. This could be used to help teens learn about game theory, or to help them develop new strategies for playing games. ai.stackexchange.com

· Catalyst ML: Catalyst ML is an AI platform that can be used to train and deploy machine learning models. This could be used to help teens learn about machine learning, or to help them develop new machine learning models for their own projects. pubs.acs.org


· Google Brain: Google Brain is a research team at Google that is working on developing new AI technologies. This could be used to help teens learn about the latest AI research, or to help them get involved in AI research themselves.www.wired.com - note summer 2023 google brain merged with deep mind


· IBM Watson: IBM Watson is an AI platform that can be used to answer questions, translate languages, and generate different creative text formats. This could be used to help teens learn about AI, or to help them develop new ways to use AI in their own lives. brainstation.io


These are just a few examples of the many AI models that are available. As AI continues to develop, there will be even more ways that AI can be used to help people.

summer 2023 chris.macrae@yahoo.co.uk writes: LLM models are changing fast. Track by asking about biggest by language, note technically summer 2023 nvidia-microsoft may be biggest model out of usa; and because of number of characters in chinese china's number 1 llm may be biggest worldwide but not necessarily most usable. There may also be two opposite main cases- an LLM intended to be the language gateway llm to web3 ie for massive use; an llm intended for strategic consulting.  Also gateway LLMs are in a global race to connect in most languages, Thus area is in flux at the moment - interesting to ask LLM their visions(If a nation intends to be eg top 10 player in AI it probably needs a gateway llm to web3 run by its biggest tech forms but with intent for public use. This also suggest that gateway LLMs may finally transform education to optimise blending and personal id of skills mastered.

NB Plugins are very important subject-and they can help more deeply customise an LLM and/or open the filed to University owned LLMs. 

As of summer 2023 we suggest reading this bard dialogue on plugins, including university and potentially UN use https://storage.ning.com/topology/rest/1.0/file/get/12149675085?profile=original  from thgis we feel these platforms are relevant for top 50 to keep watch on

LaMDA 1 of Stanford is one of the most well-known "University" LLMs, and it has been shown to be capable of performing a wide range of tasks. It is also being developed by Google AI, which is one of the leading research labs in the field of artificial intelligence..

  • Jurassic-1 Jumbo: Jurassic-1 Jumbo is an LLM developed by the Allen Institute for Artificial Intelligence. It is one of the largest LLMs, and it is trained on a massive dataset of text and code.
  • WuDao 2.0: WuDao 2.0 is an LLM developed by the Beijing Academy of Artificial Intelligence. It is one of the most powerful LLMs, and it is capable of performing a wide range of tasks.

addendum -currently i suggest fei-fei li is taking on responsibility of most connected full time educator in good ai across dozens of disciplines as well as being core mentor to white house and congress - see eg ai.gov and nairr- so i double-checked which platforms she and stanfird teams us, with bard's reply as follows

 can you tell ,me which of all of above are connected with fei-fei or stanford- and indeed which are the 5 ai model names her teams are currently spending most time on?

Sure, here are the AI models from the list above that are connected with Fei-Fei Li or Stanford:

· ImageNet: ImageNet is a large-scale image database that was created by Fei-Fei Li and her team at Stanford. ImageNet is used to train many of the AI models on this list, including GPT-3, DALL-E 2, and Bard.

Opens in a new windowwww.analyticsvidhya.com

ImageNet AI model

· Catalyst ML: Catalyst ML is an AI platform that was developed by Fei-Fei Li and her team at Stanford. Catalyst ML is used to train and deploy machine learning models, and it is available for free to students and researchers.

Opens in a new windowpubs.acs.org

Catalyst ML AI model

· Stanford Human-Centered AI Institute: The Stanford Human-Centered AI Institute is a research institute at Stanford that is led by Fei-Fei Li. The institute is focused on developing AI that is beneficial to people, and it is working on a number of AI models that are designed to help people in a variety of ways.

The five AI models that Fei-Fei Li's teams are currently spending the most time on are:

· ImageNet-R: ImageNet-R is a new version of ImageNet that is designed to be more representative of the world's population. ImageNet-R is being used to train AI models that are less biased and more likely to be helpful to people from all backgrounds.

· Bard: Bard is a large language model that is still under development, but it has already learned to perform many kinds of tasks. Bard is being used to develop new ways for people to interact with computers, and it is also being used to generate creative content.

· Catalyst ML: Catalyst ML is an AI platform that is being used to train and deploy machine learning models. Catalyst ML is being used to develop new AI models for a variety of applications, including healthcare, education, and transportation.

· Stanford Human-Centered AI Institute: The Stanford Human-Centered AI Institute is working on a number of AI models that are designed to help people in a variety of ways. These models include AI models for education, healthcare, and transportation.

.advisory boards\
UN aka aigood sdg world

the National Science Foundation (NSF) has been given a lot of responsibility for developing and funding US national AI models. The NSF has announced a number of initiatives in this area, including:

· The AI Institute for Fundamental Science: This institute will bring together researchers from a variety of disciplines to study the fundamental principles of AI.news.mit.edu

· The AI Institute for Robust Intelligence: This institute will develop AI models that are more robust and less likely to make mistakes.  www.robustintelligence.com

· The AI Institute for Technology and Society: This institute will study the social and ethical implications of AI. srinstitute.utoronto.ca

The NSF is also working with other organizations, such as Stanford University, to develop AI models. It is likely that some of the AI models that you mentioned, such as ImageNet-R, Bard, and Catalyst ML, will be used in these projects.

Here are some of the other AI models that the NSF may be cooperating or plugging into:

· AI21 Labs: AI21 Labs is a company that is developing AI models for a variety of applications, including healthcare, education, and transportation.www.ai21.com· OpenAI:  windowfourweekmba.com


· Google AI: Google AI is a research division at Google that is developing AI models for a variety of applications, including search, machine translation, and self-driving cars.ai.google


t

x we asked bard to help list some disciplines that are transfrming their oractice with AI

Here are some disciplines that are now known to have made a serious start to connecting their human intelligence with what regenerative AI can assist on:

· Computer science: Computer science is the field of study that deals with the theory, design, and implementation of computer systems. AI is a subfield of computer science, and it has been used to make significant advances in a variety of areas, including image recognition, natural language processing, and machine learning. www.mastersportal.com

· Healthcare: Healthcare is another field that has benefited greatly from AI. AI is being used to develop new diagnostic tools, to improve the accuracy of medical treatments, and to personalize patient care. www.statista.com

· Finance: AI is also being used in finance to improve trading strategies, to predict market movements, and to detect fraud.extension.harvard.edu


· Law: AI is being used in law to help lawyers with tasks such as legal research, document review, and case management.englishcomplit.unc.edu

· Education: AI is being used in education to personalize learning, to provide feedback to students, and to assess student progress. habitatbroward.org

· Engineering: AI is being used in engineering to design new products, to optimize manufacturing processes, and to improve the safety of infrastructure.www.livescience.com

· Business: AI is being used in business to improve customer service, to optimize supply chains, and to make better decisions.online.hbs.edu

These are just a few of the many disciplines that are now using AI. As AI continues to develop, it is likely that even more disciplines will start to use it.


No comments:

Post a Comment