Mediaboss Marketing

In the past few years, we have witnessed the gradual mainstreaming of large language models and the research of AI applications in the B2B industry. Despite the tremendous technological advancements made, we are still in the early stages of generative AI applications for B2B use cases. So far, the vast majority of generative AI applications have been creating new content based on a set of instructions.

Generative AI

To analyze generative AI, we first need to differentiate between B2C and B2B applications. When we, as consumers, apply generative AI, our goal is oriented towards playing, entertainment, and sharing. In terms of entertainment, quality and accuracy are not always the most important aspects, but rather the ability of the AI model to generate art or music, for example, that can be shared in a Discord channel, even if it is quickly forgotten. There is usually a psychological inclination to believe that more content equals more productivity equals better, so users are often drawn to generative, automatically-created AI tools.

When it comes to B2B applications, the business goals are different. The focus here is primarily on cost-benefit evaluations of time and quality. Either we want to produce higher quality in the same amount of time or the same level of quality in less time.

People mainly use B2B applications in the workplace, where quality is more important. However, the content generated by AI today is mainly for repetitive and low-risk work, which doesn't necessarily require high quality. It has been found that generative AI is unreliable in writing opinions or arguments. When it comes to innovation and collaboration in B2B production environments, large model-generated SEO information may be useful. But if we ask it to write a detailed blog post about a new product for developers, there will be a significant need for human input to ensure accuracy and resonance with the target audience.

Essentially, in brainstorming and early stages, the first wave of generative AI was successful in more substantive writing, but ultimately, the more creativity and domain-specific knowledge required, the more human input is needed.

Even in cases where generative AI is useful for longer blog posts, the prompt must be precise. That is, the author must already have a clear understanding of the substantive concepts that represent their blog post. Then, to get good results, the author must review the AI output, iterate the prompt, and, if necessary, rewrite the entire section.

An example here is using ChatGPT to generate legal documents, where someone familiar with the legal prompt provides all the necessary clauses and ChatGPT uses those clauses to generate a draft. However, these efforts still require a professional lawyer to review it, edit the output, and produce a sample that can be signed. This is also why the cost-benefit evaluation model is disrupted in the B2B context.

Trainable GPT-AI

Gathering information, training AI robots, and improving decision-making are crucial. When it comes to understanding the world and making decisions, humans must be involved. What AI can do is help humans apply more of their brainwaves to valuable and creative work, which means that we can not only spend more time doing important work in a day but also liberate ourselves to engage in the most valuable work. This vision is almost the opposite of ChatGPT's user interface: instead of writing lengthy responses based on a concise prompt, it is better for domain experts to train AI machines to have representative AI robots in each field: GPT-AI.

GPT-AI is a decentralized web3 project developed and created independently using CHATGPT artificial intelligence. The goal of GPT-AI is to enable everyone to have and train their own AI robots, eventually forming a huge scale of AI applications, transactions and rental platforms.

For example, if you are an image processor, designer, nutritionist, fitness coach or a chef, you can teach your AI robot your best skills and knowledge, continuously training it, accumulating data, optimizing its data structure, making it more professional. Such AI will be the most popular presence in all industries of Web3, and you can serve other users by renting or selling AI robots, thus earning commissions for yourself. This is the huge demand value that has been released by the combination of Web3 community and AI, and the value generated after solving the demand is returned to the users who keep training GPT-AI robots.

The decentralized and distributed features of Web3 provide better support for GPT-AI. In the Web3 ecosystem, all data and applications are stored on a decentralized blockchain network, which is public, transparent, and immutable. The distributed data architecture makes it easier for GPT-AI to access and share data while ensuring data security. In addition, the smart contract function of Web3 can also provide GPT-AI with more flexible and efficient transaction and training mechanisms, making the application and sale of GPT-AI more convenient.

GPT-AI is like a blessing, with its collaborative nature and more mature and humanized professional knowledge after training, its more suitable for all the Web3 user groups than the potential threats of automation and replacing humans.

Learn more: https://gpt-ai.io/

Media ContactCompany Name: GPT-AIContact Person: NICKEmail: Send EmailCountry: United StatesWebsite: https://gpt-ai.io/

Press Release Distributed by ABNewswire.comTo view the original version on ABNewswire visit: A16Z: From Information Generative AI to Trainable GPT-AI

See more here:
A16Z: From Information Generative AI to Trainable GPT-AI ... - Benzinga

Most runners know that in order to get better, they need to train the major muscle groups that are activated during a run the quads, the glutes, the hamstrings, the calves.

But few think about training the muscles that allow them to breathe specifically the diaphragm, said Kristen Konkol, an associate teaching professor of exercise science at Syracuse University. And that may be putting runners at a disadvantage.

When runners are not efficiently using their diaphragms during exercise, theyre limiting how deeply they can inhale, which in turn limits how much oxygen is absorbed and delivered to their muscles. That ultimately affects how well those muscles function during exercise, said Nicole Hagobian, a running coach and professor of kinesiology at California Polytechnic State University.

While there is not a lot of research on how diaphragmatic (or belly) breathing directly improves running, the experts we spoke with agreed that at least in theory, it can help to increase your oxygen supply during exercise.

Heres what we know about how diaphragmatic breathing can help your run and strategies for doing it properly.

The diaphragm is alarge, dome-shapedmuscle that sits under the lungs and looks like an upside down U. When you inhale, it contracts and flattens, creating a suction-like force (sort of like a syringe) that pulls air into the lungs, said Dr. Tianshi David Wu, an assistant professor of pulmonology at the Baylor College of Medicine.

Other muscles in the chest, neck and shoulders also work to bring in more air by pulling the chest up and expanding the upper and middle regions of the lungs when we inhale, Dr. Wu said.

Some people dont use their diaphragms to their fullest potential, he said, which causes them to over-rely on these other muscles. When this happens, they dont get air deep into the lungs, which limits how much oxygen they can absorb.

Diaphragmatic breathing involves consciously using your diaphragm to take deep breaths by trying to inhale into your stomach, rather than into your chest, according to Dr. Konkol.

One way you can do it is to lie on your back with your hands on your stomach and take deep breaths in through your nose, consciously trying to force the air into your belly. As you do this, the hands on your stomach should rise. As you exhale your hands and belly should recede, Dr. Konkol said.

Dr. Hagobian teaches runners the technique by placing one hand on her belly and one on her chest and then taking quick, shallow breaths to demonstrate how the hand on her chest moves much more than the hand on her belly. She then takes a deep, diaphragmatic breath, pointing out how the hand on her belly is now the one moving more while the hand on her chest is mostly still.

To get the hang of this kind of breathing and to make it more second-nature, Dr. Konkol recommended practicing the technique for 15 to 20 minutes every day or every other day for a couple of weeks. Similar to how we train our legs, we have to train our lungs, she said.

Only a few small studies have directly looked into how diaphragmatic breathing affects exercise. In one study published in 2018, for instance, researchers found that participants with fatigued diaphragms could not exercise as intensely as usual. And a 2004 studyfound that using certain breathing techniques to trainvarious muscles involved in respiration, including the diaphragm, helped cystic fibrosis patients take deeper breaths and work out more intensely on anexercisebike. A 2006 studyfound similar results in healthy adults.

Joe Shayne, a long-distance running coach with TeamWRK, a running organization in Brooklyn, said that it can be difficult to learn an entirely new way of breathing. And pressuring yourself to do it right can cause you to tense up and make breathing deeply harder.

So herecommended practicing diaphragmatic breathing while youre calm and relaxed.

He suggested practicing diaphragmatic breathing first while youre either lying down, sitting or standing, and then trying to incorporate it into walks. Once diaphragmatic breathing starts to feel easy and natural on your walks, you can graduate to using it on longer or more vigorous walks, then to jogs and eventually to more intense runs, Mr. Shayne said.

The technique becomes more challenging to maintain as the length or intensity of the exercise increases, he said.

After you have the technique down, Dr. Hagobian and Mr. Shayne recommended figuring out a breathing pattern that works for you on runs. For example, Mr. Shayne likes to exhale every four steps. Dr. Hagobian, on the other hand, prefers to start by breathing in for three steps and out for two steps, and then breathing in for two steps and out for one step at faster paces.

Having a rhythm helps you focus on your breathing technique and keeps it from becoming erratic, Dr. Hagobian said.

As you get better at taking full diaphragmatic breaths, you should see subtle but noticeable changes to your runs, Dr. Konkol said.

You should require fewer breaths per minute and you may feel more energized all because youre getting better at supplying your body with the necessary oxygen.

Read the original here:
One Way to Improve Your Run? Diaphragmatic Breathing - The New York Times

Edgar Cervantes / Android Authority

TL;DR

Google has been accused of training its AI chatbot Bard on data from OpenAIs ChatGPT without authorization. According to The Information, Jacob Devlin, a Google AI researcher, resigned because the company scraped ChatGPT data from a website called ShareGPT.

Devlin quit after sharing concerns with Pichai, Dean, and other senior managers that the Bard team, which received assistance from Brain employees, was training its machine learning model using data from OpenAIs ChatGPT, the report states. Devlin has since joined OpenAI to work on ChatGPT.

OpenAI and Google are in direct competition in the generative AI space. Microsofts heavy investment in OpenAI and the quick succession with which it integrated GPT into its products has left Google scrambling to bring its own AI-powered chatbot Bard to market. An accusation that Google lifted ChatGPT data could damage the companys reputation, given that it has spent years fostering AI research.

Unrelated to the report, Android Authority was recently approached by an SEO agency called Loopex Digital, which claimed it had a chat with Bard wherein the AI helper said it was based on OpenAIs GPT-3 language model. However, in a later exchange, Bard changed its tune and claimed to be based on Google AIs LaMDA model. For all we know, it could have been a case of Bard giving out incorrect information, a common flaw in generative AI models. However, the opposite would mean theres some truth to the latest allegations. The firm provided us with the following screenshots of its chats with Bard.

Google has categorically denied that Bard is based on ChatGPT data. Bard is not trained on any data from ShareGPT or ChatGPT, Google spokesperson Chris Pappas told The Verge.

Read the original post:
Google accused of ripping off OpenAI's ChatGPT - Android Authority

According to the World Health Organization, an estimated 1.3 billion people in the world experience a significant disability. Living with a disability is more common than you think!

The Convention on the Rights of Persons with Disabilities states disability results from the interaction between persons with impairments and attitudinal and environmental barriers that hinder their full and effective participation in society on an equal basis with others.

The inclusion of persons with disabilities is a key objective and a core area of the United Nations Disability Inclusion Strategy. It provides a foundation for sustainable and transformative change towards inclusion throughout all pillars of the organizations work, and highlights the need to strengthen system-wide accessibility and mainstream disability in all plans and efforts.

In light of this, UNSSC has taken various actions to improve the digital accessibility of its online learning content, and to create an inclusive environment. Digital accessibility refers to designing and developing online learning materials in a way that makes them perceivable, operable, understandable, and robust; ensuring they are usable by all learners. It is crucial to prioritize accessibility in online education to ensure every learner, irrespective of their personal capabilities, has equal access to learning opportunities.

Being mindful of that: how can we, as facilitators/trainers, ensure that all learners benefit from online learning offerings?

Our data training programmes feature a multidisciplinary curriculum that draws on insights from the social sciences, statistics, andmanagement while focusing mainly on datavisualization and storytelling; the fundamentals of descriptive and predictive analytics.They have been designed to better equip UN staff to formulate problem statements for data-informed solutions,use data visualization and storytellingdesign principles to communicate key findings, and build basic predictive models.

Data training material introduces learners to a variety of charts, formulas, tables, algorithms, and data processes. For instance, in our data analysis lessons, we present statistical models and techniquesfor analysis once the data has been collected and cleaned.In our data visualization modules, we explore how to select the right visualization for a data research question, and share best practices and lessons learned . In the predictive analytics sections, we switch to the features needed to explore statistical algorithms and machine learning techniques to identify the likelihood of future outcomes and applications in the UN context.

Considering the complexity of the content and the amount of visuals displayed, ensuring digital accessibility is crucial , as it enables all learners to understand the metrics and clarify their findings with appropriate explanations. In the past months, we have been working to improve the accessibility aspect of our learning products to offer an inclusive learning experience for our participants regardless of their abilities.

Below are some key aspects that we found particularly important to address to ensure an inclusive instructional design.

Audio, transcripts, visual descriptions and closed captions for videos

We provide lots of videos and interactive presentations in the modules because motion pictures reinforce and make learning easier. . We designed our videos in a way that fosters engagement by highlighting key points or asking questions to encourage learners to pause and reflect.

However, it is essential to think about those with hearing impairment, blindness, and non-native speakers. For every video and presentation, we provided a voice-over, closed captions, and time-synchronized text. We also inserted transcripts to offer a written version of the audio content and descriptions to explain all visual elements displayed in the video.

Creating visual descriptions that are both illustrative and concise is a challenge. We recognized the importance of providing accurate descriptions, but it can be difficult to understand which details are crucial to convey. We have invested considerable time and effort to develop informative and clear descriptions that work effectively for all learners.

Colour and Contrast

Colour is a powerful design tool because over time they can easily be associated with brands, themes, or emotions. However, relying on colour only is not enough if you have a visually impared or colour-blind audience. It should not be the only indicator for interactive elements. It is important to provide the right contrast between the content and the background so that the text can be easily read.

In this regard, we ensured that our visual elements respected these parameters by using a colour blindness simulator. We provided guidance in our lessons on how to identify when a data visualization does not use accessible colours.

Alternative text to images

Alt text is a brief description added to images and other visual content on web pages to ensure accessibility for the visually impaired and improve search engine optimization. It is intended for screen readers and is visible when images fail to load due to network issues. We added alt text to all images so that screen readers can read the descriptions for learners.

Describing complex images, especially those that serve as learning exercises, can be challenging. We chose to provide enough information for the users who cannot see the image, while also ensuring the information is concise and does not reveal the answer or complete the task on their behalf. To address this challenge, our team regularly discussed complex images and tried to empathize with impaired people to guarantee accurate and effective alt-text descriptions.

Consistency in design, writing style, fonts, and URL

Coherence in design by using a consistent layout and formatting for all materials is a must for us. It helps learners navigate the platform easily, reduces confusion, and improves the learning experience.

We chose accessible fonts and provided descriptive text for all URLs. We included keyboard instructions in lesson introductions, enabling users who may have difficulty using a mouse or touchpad to navigate the platform using only their keyboard.

Despite the challenges faced, we were thrilled to start this path! Our team is committed to creating an inclusive and accessible learning environment. We will continuously strive to improve our course content to ensure that it is accessible to all our learners.

Accessibility not only allows full and effective participation and inclusion in society, but also ensures a better learning experience for all. Universal design and compliance with accessibility standards provides an opportunity to create a learning environment where all users have the possibility to understand, navigate, and interact more efficiently.

Excerpt from:
Ensuring inclusive e-learning: reflections from UNSSCs data ... - United Nations System Staff College

An essential part of hanging onto the best and brightest stars on your team is making sure everyone feels fulfilled and like they are improving.

However, newsrooms are busy places and it can be hard to prioritise this amongst all the other fires to put out. Here are some ideas you can use to help gain, train and retain the most talented journalists.

During the pandemic, CNN started up its e-learning platform CNN Academy. It realised that as the world shifted to digital, it needed to follow suit with its training arm.

The Academy partners with universities and media entities around the world, where participants enrol for online training, live masterclasses with CNN pros and, more recently, a simulated newsroom experience.

The e-learning platform has a short-term, ten-course syllabus that yields a CNN qualification on completion. Those courses cover the staples of journalism ethics, breaking news on social media, and an array of storytelling techniques. The essentials to work in the frenzy of 24/7 news.

Just before Christmas 2023, CNN trialled a simulation of a rolling breaking news story, where 88 participants took part in a game over five days. Each day, the story moved on, and journalists had to chase new information, grill mock press officers, navigate a custom-made social media platform and come up with new ways to report the story.

It took place in Abu Dhabi, in the UAE, and it was designed by gaming experts as well as CNN journalists, mirroring the fast-paced dynamic of the newsroom in a "safe to fail" environment. As in, where mistakes would not prove dire or costly.

Six teams went to a final day round with a full day of editing and producing with a CNN editor, with one team coming out as a winner.

Director of CNN Academy, Alireza Haji Hosseini, said on the Journalism.co.uk podcast that this helps participants discover what their best attributes are, the true demands of the job and, ultimately, what career path to pursue.

Of the 250 academy graduates, one now works with CNN full-time, whereas others have gained internships at the broadcaster, returned to their old jobs upskilled, or even launched their own YouTube channels.

"It gives us an opportunity to spot and recruit talent from around the world," says Haji Hosseini.

"It also means that people go out into the world and do amazing things with the CNN ethos and guiding principles providing them with a path forward when it comes to factual, effective storytelling."

The goal of developing talent should not always be retaining talent, argues Amy Lewin, editor of the European startup news website Sifted, on the Journalism.co.uk podcast. In fact, it can be beneficial to be realistic about helping journalists move on in their careers.

"We know people arent going to be at Sifted their whole career. Lots of our journalists are 28, it would be madness to assume they would," she says.

"Plus, wed like to be able to say previous people who have worked here - and this has happened - now works with Reuters in Mexico. That probably helps us attract more talent in the future.

"Its being realistic about people not being here forever, but if we can get two to four amazing years out of them, and help them get a really good start or next step in their career - thats beneficial for both of us."

To that end, Lewin has been inspired by the clever startups Sifted reports on. Well-known consumer tech companies like Monzo realise that their best employees perform better when they are unshackled from the burdens of leadership. The equivalent for newsrooms is not to steer star reporters into the usual ascension of editor roles.

Instead, Sifted has a 'progression framework' to constantly monitor their skills and keep them developing where they need to. The framework lives in a shared document for the whole editorial team to see what others need help with.

Currently, there are three tracks for editorial assistants, reporters and senior reporters. It is like a school report on six key skills: writing, ideas, network and sources, sector expertise, responsibility and behaviours. It is developing another track for editors.

This shows the areas where staff are thriving, where they need improvement, how to get to the next level, and how long it may take to get a promotion.

The framework is pulled up for biannual and annual reviews, which provide agreed goals for salary increase but there are no quantitative targets for the performance of their articles.

Lewin has one-to-one meetings to ensure progress is constantly being made and can assign custom tasks, like specific original stories to dig into, or ways to widen networks. She says the framework must be reviewed regularly to account for new team members and editors should be flexible where possible.

A busy editor might not have time to spare. That is where dedicated roles in training can be useful.

Jon Yarker is an experienced financial journalist working for business content agency Rhotic Media. That was until he was given a new position as head of editorial development at the start of the year.

It is a bridge role that ensures that multiple departments can work together on a shared objective. In this case, HR, editorial and senior management are all plugged into editorial training.

It is now built into his job description and time management to oversee training, mentor graduates, design internship programmes and manage degree apprenticeships.

Yarker was speaking on the Journalism.co.uk podcast about the challenges of getting adjusted to the blurring lines between editorial and corporate financial writing.

It is similar in that there are one-to-one meetings with staff, where he can introduce actions to improve their performance, like getting them to more industry events or collaborating with different colleagues. It is a role where he is looking for signs of improvement, like whether the copy is filed quicker and they can act more independently.

"This is as vocational as it gets," says Yarker. He wants to capitalise on the enthusiasm of young reporters which is seldom explored by newsrooms that do not have the capacity or proper vision for work experience students.

"You're brimming with motivation and they get you doing something arbitrary and harmless.

"We're not going to give them the keys to everything and set them up on the expensive accounts. But we want to get them stuck in, sleeves rolled up and ready to go."

Rhotic now has two offices in London and Essex but is offering remote work and flexibility around studies to appeal to a broader remit of young journalists. As Yarker says: "if they can log on, they can learn."

There is nothing quite like hands-on training from media industry experts. If you are looking to send reporters out for a day of training, look no further than our upcoming Newsrewired conference on 23 May 2023.

We have workshops on hostile environment training with VR and generating innovative ideas, plus plenty of panel discussions on live blogging, artificial intelligence and more. Check out the full agenda here and be sure to grab your tickets here.

Alternatively, check out our range of open training sessions on mobile journalism, SEO, media law, interviewing and much more.

If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).

See more here:
Innovative ideas to gain, train and retain journalistic talent - Journalism.co.uk