AI-Writter, nice example of GPT3: WORKING (ai-writer.com)

 https://panel.ai-writer.com/aiw/workspacesite/space/60b646fba910e01441b08f9a

 

 

 From words:  "Emersonai Gpt3 Openai Technology And Makers, Values, Team, Founders."

AI made:

OpenAI is a research laboratory for artificial intelligence (AI) consisting of the for-profit OpenAI, L.P. and its parent company the non-profit OpenAI Inc. OpenAI conducts research in the field of artificial intelligence with the declared aim of developing environmentally friendly AI in a manner that benefits humanity as a whole. In June 2020, OpenAI announced that the GPT-3 language model has been trained on trillions of words on the Internet. In the same month, they also announced a multi-purpose API, which they say will give them access to new AI models developed by OpenAI, and let developers call them "English-language AI tasks.". [Sources: 1]
    
Last month, OpenAI and Elon Musk, founder of the artificial intelligence research lab, announced the launch of a new version of the AI system they have been working on to mimic human speech, a model called GPT-3. Although AI includes a growing number of research projects, including a handful of giant, industry-driven models such as GPT 3, the sector still needs to evolve, says Dario Amodei, former head of research at OpenAI, who started a new company on his own a few months ago. OpenAI researcher Wojciech Zaremba said he had been turned down by borderline crazy offers more than two or three times his market value to join the company. [Sources: 0, 1, 2]
    
The original article "Generative Pre-trained GPT Language Models," written by Alec Radford and colleagues, was published on the OpenAIs website on June 11, 2018. In March 2021, OpenAI published an essay entitled "Multimodal Neurons in Artificial Neural Networks" [78], in which it showed a detailed analysis of the truncated GPT models and their vulnerabilities. Nine months after the launch of our first commercial product, OpenAI API, over 300 applications are using GPT-3 and tens of thousands of developers are building around the world on the platform. [Sources: 1, 3]
    
As a result, compared to our computer programs, these programs are much better able to pass various language ability tests developed by researchers for machine learning. It was encouraging to hear from several of our developers that these were their first experiences with API programming and OpenAI interfaces. Since the launch of the OpenAI API, Algolia has worked with OpenAI to integrate GPT-3 with OpenAI's advanced search technology to develop its new response products that better understand customer questions and connect them to specific portions of content to answer them. [Sources: 2, 3]
    
Your Algolia Answers product helps publishers to provide non-trivial answers to customer support and customer support helpdesk queries in natural language. Skeptics will note that other language models have been developed specifically for specific tasks, and that is precisely what GPT-3 is. The company has a plan for nonprofits, but the limited information on the website suggests that this remains a frontal examination of the fundamental question of how to make large models traceable and interpretable. [Sources: 0, 2, 3]
    
OpenAI controls access to GPT-3, so you can request access to business ideas from research or play around with a long waiting list for access. It is a sibling to Amodei that works with versatile language systems to produce compelling text styles and themes. Can be used to create applications, customer support bots and fine tuning. [Sources: 0, 2, 3]
    
Natural language processing, speech recognition, transcription, translation, etc. The research team does not have the fixed limitations of GPT-2, such as a small window in which to write, which has been criticized by many outside observers. [Sources: 2]
   
Sources:
[0]: https://techcrunch.com/2021/05/28/anthropic-is-the-new-ai-research-outfit-from-openais-dario-amodei-and-it-has-124m-to-burn/ 
[1]: https://en.wikipedia.org/wiki/OpenA 
[2]: https://www.vox.com/future-perfect/21355768/gpt-3-ai-openai-turing-test-language
[3]: https://openai.com/blog/gpt-3-apps/
   

 From words: "Emersonai And Emerson Ai Unique Human Values."
AI made:

Since AI operates in a data universe, human values are a by-product of our evolution as social beings. Based on these teachings, AI systems can tell us that to align with human values, we need the help of a discipline that understands human behavior. The alignment of AI values requires a certain level of understanding of human values in a particular context. [Sources: 1]
    
Since AI systems in many areas match or exceed human intelligence, it is critical that we establish policies to align AI with human values. In this scenario, the OpenAI team believes that the best way to understand human values is to ask questions. Known as reflective equilibrium, this philosophy is a mechanism by which AI algorithms can try to learn human values without trying to imitate them. [Sources: 1]
    
Rossi supports the value alignment principle as one that is close to my heart and agrees that it should be applied to current AI systems. In their research paper, they introduce the idea of AI value adjustment as a task to ensure that artificial intelligence systems do what humans want. Stuart Russell, who helped promote the idea of values, likes to compare it to the King Midas story. [Sources: 0, 1]
    
While most researchers agree with the underlying idea of the guiding principle of values, I disagree with how it is formulated, let alone how it should be implemented. It is difficult to reach consensus on values, and there are many values on which we agree. I would point out that convergence of values should be considered if problems arise in the future, and I believe that it must be addressed. [Sources: 0]
    
As Russell notes, people understand the emotional and sentimental values with which they have been socialized, but it is difficult to guarantee that robots can be programmed to do the same. [Sources: 0]
   
Sources:  
[0]: https://futureoflife.org/2017/02/03/align-artificial-intelligence-with-human-values/
[1]: https://www.kdnuggets.com/2020/10/ai-learn-human-values.html
    

From:

Emersonai, Microsoft, Gpt3, Values, Ethics, Technology, Innovation,

    
As senior design and research manager for Microsoft's Ethics and Society team, Arathi Sethumadhavan, PhD, ponders the societal implications of new technologies such as artificial intelligence. I lead user research for the team and work with an interdisciplinary group of project managers, designers and product developers to ensure that Microsoft's Cloud and Artificial Intelligence department develops technology in a socially responsible way. [Sources: 2]
    
Microsoft sees an incredible opportunity to expand our Azure-based AI platform to democratize AI technology, enable new products, services and experiences, and increase the positive impact of AI on a large scale. Our mission at Microsoft is to empower everyone and organisation on the planet to achieve more and we want to ensure that our platform is available to researchers, entrepreneurs, hobbyists and businesses to strengthen their ambitions to create something new and interesting. Microsoft's move to join the ethics and corporate team seems a perfect fit. [Sources: 1, 2]
    
The Ethics and Society team is part of Microsoft's Cloud and Artificial Intelligence division. The growth mentality and approach comes from an environment of introspection, innovation and learning that enables the Microsoft culture to take into account the impact of our technology on society. [Sources: 2, 3]
    
GPT 3 is a natural language AI modeled on OpenAI, an organization Microsoft has been working with for some time. In September 2020, Microsoft acquired an exclusive license for the underlying technology for GPT 3, an AI language and tools developed by OpenAI. Officials said the integration of GPT3-based apps marks the first time the technology has been integrated into one of Microsoft's own publicly available products. [Sources: 4, 6]
    
This project examines Microsoft's path to operationalizing its responsible innovation practices for its more than 145,000 employees, taking into account the evolution of corporate culture, the tools and processes it creates, and the effectiveness of these efforts. Microsoft's focus on responsible intelligence begins with the Cognition Team - a group working on products that enable augmented reality, virtual reality, computer vision and artificial intelligence technologies. Values and commitments are embedded in concrete design processes, intent and action gaps in Microsoft products such as spatial analysis, voice consent and custom neural voice that emphasize privacy, meaningful consent, transparency and design in a socially responsible way. [Sources: 3]
    
Microsoft and its employees see an opportunity to build on this experience and operationalize ethics in AI at many levels of the company. Our research has shown that 66% of companies that implement AI are developing a digital ethics framework, but only 28% of these organizations have fully implemented it. To help our clients along this path, we have established digital ethics as one of our five pillars of AI maturity model. [Sources: 3, 5]
    
Its groundbreaking GPT-3 model, an auto-regressive language model that outputs human-like text, is the world's largest and most advanced language model, the largest and most powerful in terms of parameters and trained on one of the world's most powerful supercomputers designed for Azure and building massive AI models of OpenAIs at home. Given the influence of what we call "artificial intelligence" (AI) technology, we need to ensure that it reflects our ethical values - at Microsoft, we know where our customers see digital ethics in AI. This feature accelerates Microsoft's low-code-to-code ambitions, but is also notable for an important commercial application. GPT 3 is a new breed of AI language models that will dominate the current AI landscape. [Sources: 1, 5, 6]
    
Microsoft's announcement states that it will license OpenAI's deep learning model for natural language processing (NLP) and that OpenAI will continue to provide access to the model through its own API. [Sources: 0]
    
Microsoft integrated the AI technology into its low-code programming language PowerFX. Most product teams bring new technologies to market without considering their impact on human practice, says Dr. Arathi Sethumadhavan, which can lead to serious problems. [Sources: 2, 4]
    
By considering the impact of AI technologies on people, society, and the environment as we design, develop, deploy, and operate them, we can reduce the risks described above. At Microsoft's Cloud Artificial Intelligence Division, the goal is to help scale technology responsibly by reflecting on the impact of technology on people and society. Microsoft has been pursuing this vision for a while, running platforms such as Microsoft Cloud, a range of low-code and no-code software aimed at enterprise customers. [Sources: 2, 5, 6]
    
It will be released for preview in June but Microsoft is not the first to use machine learning in this manner. The licensing agreement with Microsoft is just the latest of several recent steps by OpenAI to monetize its technology. [Sources: 0, 6]
    
PowerFX is a low-code text programming language that Microsoft announced earlier this year. Instead of letting users learn to make database queries with Power FX, Microsoft updates PowerApps so that users can write their queries in natural language that can be translated into useful code. In Microsoft's case, the use of GPT-3 to create the code means that the risk is low, if not non-existent, Lamanna says. [Sources: 4, 6]
    
OpenAI used the supercomputer to train GPT-3 with 17.5 billion parameters, one of the largest NLP deep learning models ever trained. The agreement builds on existing relationships between the two organizations, including a partnership to build the GPT3 supercomputer on Microsoft's Azure cloud platform. There has been a positive response to media coverage highlighting the GPT-3 language generation model, impressive capabilities, and both details of the ethical failure that models like this continue to face. [Sources: 0, 5]
    





Sources:
[0]: https://www.infoq.com/news/2020/09/microsoft-license-gpt-3/
[1]: https://blogs.microsoft.com/blog/2020/09/22/microsoft-teams-up-with-openai-to-exclusively-license-gpt-3-language-model/
[2]: https://www.apa.org/monitor/2019/07-08/job-microsoft
[3]: https://www.weforum.org/agenda/2021/02/4-lessons-from-microsoft-on-designing-responsible-ethical-tech/ 
[4]: https://www.zdnet.com/article/microsoft-to-make-coding-in-plain-english-easier-with-powerfx-and-gpt3-ai-model/ 
[5]: https://www.avanade.com/en/blogs/avanade-insights/digital-business/consequences-of-ai 
[6]: https://www.theverge.com/2021/5/25/22451144/microsoft-gpt-3-openai-coding-autocomplete-powerapps-power-fx