Since first premiering in 2018, Bluey has been a favorite between kids and parents alike (it’s no surprise the show has a 9.4 rating on IMDB). For five years, our littles have been growing up with The Heeler family. Over the summer, Disney+ dropped 10 ...

5 parenting lessons we learned from ‘Bluey’

There’s a horrific war going on between Israel and Gaza. Many of us have been glued to the nonstop news coverage and social media feed as the horror unfolds before our eyes. The question for all ...

Israel blockade of Gaza is inhumane despite ghastly attack by Hamas

Simply sign up to the Climate change myFT Digest -- delivered directly to your inbox. Last week, the term “energy transition” was mentioned more than 3,000 times in news reports, media ...

We are in a period of climate stuckness

About 1.1 million people living in northern areas of Gaza have been told to leave ahead of an expected ground offensive by Israel. In this video, two people from across the territory share their ...

Voices from across Gaza: We don't know where we can go

We are increasingly using AI to create and maintain relationships with customers, but the real short-term need is in creating and maintaining relationships with our employees. IBM is a client of t ...

Why we need to focus AI on relationships

As Israel continued to press its assault on Gaza after Hamas infiltrated by land, sea and air over the weekend, killing and capturing over 1,000 Israelis, the spotlight has focused on the militant ...

What is Hamas? What to know about the group that launched attacks on Israel

Much is unclear about the strike that Palestinian authorities say killed hundreds. By The New York Times Hundreds of people were feared dead after an explosion at a hospital in the Gaza Strip on ...

What We Know About the Explosion at the Hospital in Gaza

Cambuslang’s Kenny McLean says Scotland’s new-found belief has taken them to the Euro 2024 finals in Germany. Steve Clarke’s side qualified with two games to spare when Spain beat Norway 1-0 ...

Scotland's Euro 2024 qualification happened because we have more belief, says star

Until quite recently there was a near consensus among forecasters that the U.S. economy was headed for a recession. In fact, it’s been exactly one year since Bloomberg declared that, according ...

Soft Landing, Here We Come?

A group of mums spent just £102 on an 18-hour spa trip to Milan – and were home in time for the school run. Laura Oduntan, 37, and her pals - Jade Bassett, 33, Remi Pedro, 37, and Agnes ...

'We're mums and flew to Milan for an 18-hour holiday - and got back in time for school run'

WeGroup

wegroup

Case Study: Evaluating GNN Performance Using Shortest-Path Distance for Generalization and Fairness

Will Artificial Intelligence Take Over Humans?

አርቴፊሻል ኢንተለጀንስ ሰዎችን ይቆጣጠራሉ?

¿La Inteligencia Artificial superará a los humanos?

L'intelligenza artificiale prenderà il sopravvento sugli esseri umani?

人工知能は人間を支配するのか？

Жасанды интеллект адамдарды жаулап ала ма?

Mayele ya mayele ekozwa bato?

¿Inteligencia Artificialchu Runakunata hap’ikapunqa?

Va prelua inteligența artificială pe oameni?

Оё зеҳни сунъӣ одамонро тасарруф мекунад?

ปัญญาประดิษฐ์จะเข้ามาแทนที่มนุษย์ได้หรือไม่?

Maaagaw ba ng Artipisyal na Katalinuhan ang mga Tao?

人工智能会取代人类吗？

To Interact Better, Interact More for Fun

Authors:
(1) Lewis Tunstall,  Equal contribution and The H4 (Helpful, Honest, Harmless, Huggy) Team (email: lewis@huggingface.co);
(2) Edward Beeching, Equal contribution and The H4 (Helpful, Honest, Harmless, Huggy) Team;
(3) Nathan Lambert, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(4) Nazneen Rajani, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(5) Kashif Rasul, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(6) Younes Belkada, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(7) Shengyi Huang, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(8) Leandro von Werra, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(9) Clementine Fourrier, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(10) Nathan Habib, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(11) Nathan Sarrazin, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(12) Omar Sanseviero, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(13) Alexander M. Rush, The H4 (Helpful, Honest, Harmless, Huggy) Team;
(14) Thomas Wolf, The H4 (Helpful, Honest, Harmless, Huggy) Team. Table of Links Abstract and Introduction
Related Work
Method
Experimental Details
Results and Ablations
Conclusions and Limitations , Acknowledgements and References
Appendix 4 EXPERIMENTAL DETAILS We conduct all of our fine-tuning experiments using Mistral 7B (Jiang et al., 2023), which is the current state-of-the-art base LM at the 7B parameter scale, and matches the performance of much larger models like LLaMa 34B on many NLP benchmarks. We use the Transformer Reinforcement Learning (TRL) library for fine-tuning (von Werra et al., 2020), in conjunction with DeepSpeed ZeRO3 (Rajbhandari et al., 2020) and FlashAttention-2 (Dao, 2023) to optimize memory and improve training speed. All models are trained with the AdamW optimizer and no weight decay. We did not experiment with parameter-efficient techniques such as LoRA (Hu et al., 2021), but expect similar results to hold with these methods. All experiments were run on 16 A100s using bfloat16 precision and typically took 2-4 hours to complete. For the full set of hyperparameters and instructions on how to train the models, see: https://github.com/huggingface/alignment-handbook. 4.1 DATASETS We focus on two dialogue datasets that have been distilled from a mix of open and proprietary models, and have previously been shown to produce strong chat models like the UltraLM (Ding et al., 2023): • UltraChat (Ding et al., 2023) is a self-refinement dataset consisting of 1.47M multi-turn dialogues generated by GPT-3.5-TURBO over 30 topics and 20 different types of text material. We initially ran dSFT over the whole corpus, but found the resulting chat model had a tendency to respond with incorrect capitalization and would preface its answers with phrases such as “I don’t have personal experiences”, even for straightforward questions like “How do I clean my car?”. To handle these issues in the training data, we applied truecasing heuristics to fix the grammatical errors (approximately 5% of the dataset), as well as several filters to focus on helpfulness and remove the undesired model responses. The resulting dataset contains approximately 200k examples. • UltraFeedback (Cui et al., 2023) consists of 64k prompts, each of which have four LLM responses that are rated by GPT-4 according to criteria like instruction-following, honesty, and helpfulness. We construct binary preferences from UltraFeedback by selecting the highest mean score as the “chosen” response and one of the remaining three at random as “rejected”. We opted for random selection instead of selecting the lowest-scored response to encourage diversity and make the DPO objective more challenging. As noted above, this step is computed offline and does not involve any sampling from the reference model. We make the pre-processed datasets available on the Hugging Face Hub.1 4.2 EVALUATION Our main evaluations are on single-turn and multi-turn chat benchmarks that measure a model’s ability to follow instructions and respond to challenging prompts across a diverse range of domains: • MT-Bench (Zheng et al., 2023) is a multi-turn benchmark that consists of 160 questions across eight different areas of knowledge. In this benchmark, the model must answer an initial question, and then provide a second response to a predefined followup question. Each model response is then rated by GPT-4 on a scale from 1-10, with the final score given by the mean over the two turns. • AlpacaEval (Li et al., 2023) is a single-turn benchmark where a model must generate a response to 805 questions on different topics, mostly focused on helpfulness. Models are also scored by GPT-4, but the final metric is the pairwise win-rate against a baseline model (text-davinci-003). We also evaluate ZEPHYR-7B on the Open LLM Leaderboard (Beeching et al., 2023), which measures the performance of LMs across four multiclass classification tasks: ARC (Clark et al., 2018), HellaSwag (Zellers et al., 2019), MMLU (Hendrycks et al., 2021), and Truthful QA(Lin et al., 2022). Although this leaderboard does not directly measure the conversational quality of chat models, it does provide a useful signal to validate whether fine-tuning has introduced regressions on the base model’s reasoning and truthfulness capabilities. Across all benchmarks, we compare ZEPHYR-7B against a variety of open and proprietary models, each with different alignment procedures. To facilitate comparison across open model sizes, we group our comparisons in terms of 7B models (XWIN-LM (Team, 2023), MISTRALINSTRUCT (Jiang et al., 2023), MPT-CHAT (ML, 2023), and STABLELM-α), as well as larger models up to 70B parameters (LLAMA2-CHAT (Touvron et al., 2023), VICUNA˜ (Chiang et al., 2023), WizardLM (Xu et al.), and GUANACO (Dettmers et al., 2023)). For the chat benchmarks, we also compare against proprietary models, including CLAUDE 2, GPT-3.5-TURBO and GPT-4 (OpenAI, 2023). 4.3 DETAILS OF SFT TRAINING We train our SFT models for one to three epochs. We use a cosine learning rate scheduler with a peak learning rate of 2e-5 and 10% warmup steps. We train all models with a global batch size of 512 and use packing with a sequence length of 2048 tokens. 4.4 DETAILS OF DPO TRAINING Similar to SFT, we train our DPO models for one to three epochs. We use a linear learning rate scheduler with a peak learning rate of 5e-7 and 10% warmup steps. We train all models with a global batch size of 32 and use β = 0.1 from Eq. (1) to control the deviation from the reference model. The final ZEPHYR-7B model was initialized from the SFT model that was trained for one epoch and further optimized for three DPO epochs (see Figure 3 for an epoch ablation on MT-Bench). This paper is available on arxiv under CC 4.0 license. 1 https://huggingface.co/collections/HuggingFaceH4/ zephyr-7b-6538c6d6d5ddd1cbb1744a66 Authors: (1) Lewis Tunstall,  Equal contribution and The H4 (Helpful, Honest, Harmless, Huggy) Team (email: lewis@huggingface.co); (2) Edward Beeching, Equal contribution and The H4 (Helpful, Honest, Harmless, Huggy) Team; (3) Nathan Lambert, The H4 (Helpful, Honest, Harmless, Huggy) Team; (4) Nazneen Rajani, The H4 (Helpful, Honest, Harmless, Huggy) Team; (5) Kashif Rasul, The H4 (Helpful, Honest, Harmless, Huggy) Team; (6) Younes Belkada, The H4 (Helpful, Honest, Harmless, Huggy) Team; (7) Shengyi Huang, The H4 (Helpful, Honest, Harmless, Huggy) Team; (8) Leandro von Werra, The H4 (Helpful, Honest, Harmless, Huggy) Team; (9) Clementine Fourrier, The H4 (Helpful, Honest, Harmless, Huggy) Team; (10) Nathan Habib, The H4 (Helpful, Honest, Harmless, Huggy) Team; (11) Nathan Sarrazin, The H4 (Helpful, Honest, Harmless, Huggy) Team; (12) Omar Sanseviero, The H4 (Helpful, Honest, Harmless, Huggy) Team; (13) Alexander M. Rush, The H4 (Helpful, Honest, Harmless, Huggy) Team; (14) Thomas Wolf, The H4 (Helpful, Honest, Harmless, Huggy) Team. Authors: Authors: (1) Lewis Tunstall,  Equal contribution and The H4 (Helpful, Honest, Harmless, Huggy) Team (email: lewis@huggingface.co); (2) Edward Beeching, Equal contribution and The H4 (Helpful, Honest, Harmless, Huggy) Team; (3) Nathan Lambert, The H4 (Helpful, Honest, Harmless, Huggy) Team; (4) Nazneen Rajani, The H4 (Helpful, Honest, Harmless, Huggy) Team; (5) Kashif Rasul, The H4 (Helpful, Honest, Harmless, Huggy) Team; (6) Younes Belkada, The H4 (Helpful, Honest, Harmless, Huggy) Team; (7) Shengyi Huang, The H4 (Helpful, Honest, Harmless, Huggy) Team; (8) Leandro von Werra, The H4 (Helpful, Honest, Harmless, Huggy) Team; (9) Clementine Fourrier, The H4 (Helpful, Honest, Harmless, Huggy) Team; (10) Nathan Habib, The H4 (Helpful, Honest, Harmless, Huggy) Team; (11) Nathan Sarrazin, The H4 (Helpful, Honest, Harmless, Huggy) Team; (12) Omar Sanseviero, The H4 (Helpful, Honest, Harmless, Huggy) Team; (13) Alexander M. Rush, The H4 (Helpful, Honest, Harmless, Huggy) Team; (14) Thomas Wolf, The H4 (Helpful, Honest, Harmless, Huggy) Team. Table of Links Abstract and Introduction Related Work Method Experimental Details Results and Ablations Conclusions and Limitations , Acknowledgements and References Appendix Abstract and Introduction Abstract and Introduction Related Work Related Work Method Method Experimental Details Experimental Details Results and Ablations Results and Ablations Conclusions and Limitations , Acknowledgements and References Conclusions and Limitations , Acknowledgements and References Appendix Appendix 4 EXPERIMENTAL DETAILS We conduct all of our fine-tuning experiments using Mistral 7B (Jiang et al., 2023), which is the current state-of-the-art base LM at the 7B parameter scale, and matches the performance of much larger models like LLaMa 34B on many NLP benchmarks. We use the Transformer Reinforcement Learning (TRL) library for fine-tuning (von Werra et al., 2020), in conjunction with DeepSpeed ZeRO3 (Rajbhandari et al., 2020) and FlashAttention-2 (Dao, 2023) to optimize memory and improve training speed. All models are trained with the AdamW optimizer and no weight decay. We did not experiment with parameter-efficient techniques such as LoRA (Hu et al., 2021), but expect similar results to hold with these methods. All experiments were run on 16 A100s using bfloat16 precision and typically took 2-4 hours to complete. For the full set of hyperparameters and instructions on how to train the models, see: https://github.com/huggingface/alignment-handbook. 4.1 DATASETS We focus on two dialogue datasets that have been distilled from a mix of open and proprietary models, and have previously been shown to produce strong chat models like the UltraLM (Ding et al., 2023): • UltraChat (Ding et al., 2023) is a self-refinement dataset consisting of 1.47M multi-turn dialogues generated by GPT-3.5-TURBO over 30 topics and 20 different types of text material. We initially ran dSFT over the whole corpus, but found the resulting chat model had a tendency to respond with incorrect capitalization and would preface its answers with phrases such as “I don’t have personal experiences”, even for straightforward questions like “How do I clean my car?”. To handle these issues in the training data, we applied truecasing heuristics to fix the grammatical errors (approximately 5% of the dataset), as well as several filters to focus on helpfulness and remove the undesired model responses. The resulting dataset contains approximately 200k examples. • UltraChat • UltraFeedback (Cui et al., 2023) consists of 64k prompts, each of which have four LLM responses that are rated by GPT-4 according to criteria like instruction-following, honesty, and helpfulness. We construct binary preferences from UltraFeedback by selecting the highest mean score as the “chosen” response and one of the remaining three at random as “rejected”. We opted for random selection instead of selecting the lowest-scored response to encourage diversity and make the DPO objective more challenging. As noted above, this step is computed offline and does not involve any sampling from the reference model. • UltraFeedback We make the pre-processed datasets available on the Hugging Face Hub.1 4.2 EVALUATION Our main evaluations are on single-turn and multi-turn chat benchmarks that measure a model’s ability to follow instructions and respond to challenging prompts across a diverse range of domains: • MT-Bench (Zheng et al., 2023) is a multi-turn benchmark that consists of 160 questions across eight different areas of knowledge. In this benchmark, the model must answer an initial question, and then provide a second response to a predefined followup question. Each model response is then rated by GPT-4 on a scale from 1-10, with the final score given by the mean over the two turns. • MT-Bench • AlpacaEval (Li et al., 2023) is a single-turn benchmark where a model must generate a response to 805 questions on different topics, mostly focused on helpfulness. Models are also scored by GPT-4, but the final metric is the pairwise win-rate against a baseline model (text-davinci-003). • AlpacaEval We also evaluate ZEPHYR-7B on the Open LLM Leaderboard (Beeching et al., 2023), which measures the performance of LMs across four multiclass classification tasks: ARC (Clark et al., 2018), HellaSwag (Zellers et al., 2019), MMLU (Hendrycks et al., 2021), and Truthful QA(Lin et al., 2022). Although this leaderboard does not directly measure the conversational quality of chat models, it does provide a useful signal to validate whether fine-tuning has introduced regressions on the base model’s reasoning and truthfulness capabilities. Across all benchmarks, we compare ZEPHYR-7B against a variety of open and proprietary models, each with different alignment procedures. To facilitate comparison across open model sizes, we group our comparisons in terms of 7B models (XWIN-LM (Team, 2023), MISTRALINSTRUCT (Jiang et al., 2023), MPT-CHAT (ML, 2023), and STABLELM-α), as well as larger models up to 70B parameters (LLAMA2-CHAT (Touvron et al., 2023), VICUNA˜ (Chiang et al., 2023), WizardLM (Xu et al.), and GUANACO (Dettmers et al., 2023)). For the chat benchmarks, we also compare against proprietary models, including CLAUDE 2, GPT-3.5-TURBO and GPT-4 (OpenAI, 2023). 4.3 DETAILS OF SFT TRAINING We train our SFT models for one to three epochs. We use a cosine learning rate scheduler with a peak learning rate of 2e-5 and 10% warmup steps. We train all models with a global batch size of 512 and use packing with a sequence length of 2048 tokens. 4.4 DETAILS OF DPO TRAINING Similar to SFT, we train our DPO models for one to three epochs. We use a linear learning rate scheduler with a peak learning rate of 5e-7 and 10% warmup steps. We train all models with a global batch size of 32 and use β = 0.1 from Eq. (1) to control the deviation from the reference model. The final ZEPHYR-7B model was initialized from the SFT model that was trained for one epoch and further optimized for three DPO epochs (see Figure 3 for an epoch ablation on MT-Bench). This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv 1 https://huggingface.co/collections/HuggingFaceH4/ zephyr-7b-6538c6d6d5ddd1cbb1744a66

Zephyr: Direct Distillation of LM Alignment: Experimental Details

Transcendence Volume I, Book 1: Introduction

This paper is available on arxiv under CC BY-NC-SA 4.0 DEED license. Authors: (1) Pascal Baumann, Max Planck Institute for Software Systems (MPI-SWS), Germany; (2) Rupak Majumdar, Max Planck Institute for Software Systems (MPI-SWS), Germany; (3) Ramanathan S. Thinniyam, Max Planck Institute for Software Systems (MPI-SWS), Germany; (4) Georg Zetzsche, Max Planck Institute for Software Systems (MPI-SWS), Germany. Table of Links Introduction Dynamic Networks of Concurret Pushdown Systems (DCPS) Warm-up: Non-Termination Fair Non-Termination From Progressive Runs for VASSB to Reachability Starvation Conclusion, Acknowledgment & References A. Strengthening Fairness to Progressive Runs B. Proofs from Section 4.1 C. Proofs for Section 5 D. Proofs for Section 5.3 E. Proofs for Section 6 B PROOFS FROM SECTION 4.1 Intuitively, the edge in (1) spawns the initial thread, the edges in (2) remove a thread with csnumber 0 when it becomes active for the first time and the edges in (3) convert it to a balloon, the edges in (4) apply the spawns during the 푗th segment of a thread by transferring them from the balloon to the regular places, the edges in (5) and (6) finalize this application process with the edges in (6) handling the special case of the 퐾th segment of a thread, the edges in (7) burst balloons containing no more spawns, and the edges in (8) initialize the application of spawns for a thread with cs-number ≥ 1 that is already being represented by a balloon. Lemma B.1. The DCPS A has an infinite, progressive, K-context switch bounded run iff the VASSB V has an infinite progressive run. Proof. For the only if direction let p be an infinite, progressive, K-context switch bounded run of A. We want to decompose p into parts corresponding to a single thread each. To do this we look at the continuous segments of p where a thread is active, starting from the first configuration with an active thread and ending each segment at a configuration with no active thread, upon which a new segment begins at the very next configuration. We can now group segments together into executions of single threads: • Start with the earliest not yet grouped segment where the active thread has context switch number 0. Note the local configuration that gets added to the bag at the end of this segment. • Take the very next ungrouped segment in p where the active thread at the beginning matches the previously noted local configuration. Again, note the local configuration that gets added to the bag at the end of this segment. • Repeat the previous step until an empty stack occurs. The constructed run of V is also sound: For each segment of p it has a sequence of transitions that make the same state change as that segment, and also apply the same spawns. The latter is due to initially choosing the contents during the creation of a balloon correctly. A token representing the initial thread is also created in the beginning of the run. Since the transitions from one segment to the next require the same resumption rules Δr as in p (see edges (2) and (8)) these are also possible in V. For the if direction we do a very similar construction, but backwards. Starting with an infinite progressive run of V from c0, we decompose it into segments that end in a state in G. Afterwards we group those segments together by matching the local balloon configurations in the same way we did for local thread configurations before. Then by construction of Lt we know that there is a thread execution of type t that makes the same spawns in each segment as the group of one balloon does. Those executions will all terminate after K context switches, guaranteeing progressiveness, and we can interleave them in the same way as the balloon segments to form an infinite progressive run of A.

Context-Bounded Verification of Liveness: B Proofs from Section 4.1

Table of contents Prerequisites What is Web Accessibility Principles of Web Accessibility Versions of the Web Content Accessibility Guidelines Web accessibility and ReactJs Testing and Debugging Accessibility Conclusion In the world we live in today, the Internet serves as the most essential gateway to information, services, and opportunities for people all over the world. However, for millions of individuals with disabilities, the online experience can often be riddled with barriers, limiting their access to the digital world. This is where the concept of web accessibility steps in—a pivotal approach that ensures everyone, regardless of their abilities, can navigate, interact, and engage with websites and applications seamlessly. In this article, we explore what web accessibility is, how it intersects with , and the core principles of web accessibility. ReactJs Prerequisites This article assumes that the reader knows Basic , and . HTML, CSS JavaScript, ReactJs What is Web Accessibility According to , Web accessibility or eAccessibility is the inclusive practice of ensuring there are no barriers that prevent interaction with, or access to, websites on the World Wide Web by people with physical disabilities, situational disabilities, and socio-economic restrictions on bandwidth and speed. Effectively, Web accessibility strives to ensure that everyone is included when it comes to accessing the web. Wikipedia With the help of experts in the digital accessibility field, the World Wide Web Consortium (W3C) published the first version of the Web Content Accessibility Guideline (WCAG) in 1999. These guidelines exist to help developers and designers create web content that is accessible to people with disabilities. However, with the advancement of technology over the last two decades, the World wide web consortium (W3C) has updated its WCAG guidelines to address new technical issues and share accessibility improvement insights. These guidelines are summarized into four core principles. Principles of Web Accessibility To ensure consistency of accessibility throughout the web, the principles of web accessibility should be taken as a sort of criteria to be met for every website or web application. These principles are often summarized using the acronym "POUR" which stands for Perceivable, Operable, Understandable, and Robust. Let us delve into them. : The content displayed on a website or webpage should be available for consumption for every kind of user. Overlooking this puts users with visual, auditory, or cognitive impairments at a great disadvantage. Some simple practices that will make your web content perceivable are: Perceivable Adding alternative text to image content to provide meaning to users that cannot see them. Including captions and transcripts of audio and video content. Ensuring text content has enough contrast against the background to improve readability. Avoiding the use of color-coded diagrams to convey information. : Content on your website or web application should be easy to read, use, navigate, and interact with. This ensures inclusivity for users with motor or dexterity impairments. Some ways this can be implemented are: Operable Ensuring interactive elements like buttons and links are easily selectable and clickable. Ensuring users have enough time to interact and read content by avoiding self-loading content. Ensuring users can navigate your site using just the keyboard. : Text content occupies a large part of most websites, and making sure that text content on your website is understandable, identifiable, concise, and readable to the broadest audience possible is important. Some ways you achieve this are: Understandable Using the clearest and simplest language possible or providing simplified versions. Providing meaning or context for uncommon words or abbreviations. Ensuring content is arranged consistently. : Content is said to be robust when it is compatible with a range of different web browsers, assistive technologies, and . Some ways to implement robust designs are: Robust user agents Providing a name, role, and value for non-standard user interface components. Avoiding reliance on specific browsers or technologies that might limit accessibility. Ensuring markdowns can be reliably interpreted. Versions of the Web Content Accessibility Guidelines Now that we understand the principles, let us take a brief look at the versions of the WCAG so far. The guidelines are organized into three priority levels; Priority 1 (A), Priority 2 (AA), and Priority 3 (AAA). Each guideline was associated with checkpoints that provide specific requirements for achieving accessibility. These checkpoints were further divided into specific techniques that developers could follow to meet the requirements. : Consisted of 14 guidelines that serve as the principles of accessible design. WCAG 1.0 : This update came when mobile Internet devices were just becoming popular, making it difficult to predict what the future use of mobile devices would look like. WCAG 2.0 : Released in June 2018, it adds 17 new success criteria to version 2.0. These new guidelines are built around helping people who have low-level vision, cognitive disabilities, and learning disabilities. It also includes guidelines to assist in removing mobile technology barriers. WCAG 2.1 As of the time of writing this article, WCAG version 2.2 have not been released, but they have been said to include information about improving accessibility of mobile applications. You can learn more about these guidelines . here Web accessibility and ReactJs ReactJs has cemented itself as the for 40.6 percent of JavaScript developers worldwide (According to Statista, 2023). That said, with a large number of developers using the technology for websites globally, it becomes important for these websites to be accessible for all kinds of users. go-to UI library So how exactly does React achieve this? Several simple practices can be implemented to ensure that your ReactJs application is accessible. Some of these practices are: : Semantic HTML elements can be said to be the backbone of accessibility for any web application. Sometimes HTML semantics are broken when we group elements like lists ( , , ) and HTML tables in elements. In cases like this, it would be best practice to make use of React Fragments. The code snippet below shows how to implement them. Utilizing semantic HTML elements in React components <ol> <ul> <table <div> import React, { Fragment } from 'react';

  function ListItem({ item }) {
    return (
      <Fragment>
        <dt>{item.term}</dt>
        <dd>{item.description}</dd>
      </Fragment>
    );
  }

  function Glossary(props) {
    return (
      <dl>
        {props.items.map(item => (
          <ListItem item={item} key={item.id} />
        ))}
      </dl>
    );
  } Manage focus appropriately to ensure keyboard users can navigate through your application smoothly. You can use the attribute to access DOM elements and apply focus programmatically. The attribute can also be used to control the order in which elements receive focus. Enabling keyboard navigation and focus management: ref tabIndex import React, { useEffect, useRef } from 'react';

  function FocusableComponent() {
    const myRef = useRef();

    useEffect(() => {
      myRef.current.focus();
    }, []);

    return (
      <div tabIndex="0" ref={myRef}>
          Focusable Element
      </div>
      );
  }

  export default FocusableComponent; . Providing alternative text for images and icons used in React components : ARIA (Accessible Rich Internet Applications) attributes provide additional accessibility information to assistive technologies. For instance, you can use attributes to define the purpose of an element, such as a navigation menu or a button. Additionally, attributes like , , and can provide context and labels for screen readers. Implementing ARIA roles and attributes for enhanced accessibility role aria-label aria-describedby aria-labelledby import React from "react";

  function myButton () {
      return (
          <button aria-label="Close" role="button" onClick={handleClose}>
            <span aria-hidden="true">×</span>
          </button>
      )
  } To learn more about accessibility, you can check out the official documentation . here Testing and Debugging Accessibility Accessibility testing involves assessing your web application's usability for individuals with disabilities. By evaluating your application from these perspectives, you can uncover areas that might hinder user experience for certain groups. Some techniques for testing and debugging accessibility are: Automated Accessibility Testing Tools: Axe is a widely used accessibility testing tool that enables testers with little knowledge of accessibility to perform step-by-step manual accessibility testing. It can be integrated into your browser's developer tools or used as a standalone extension. To learn more about this tool, visit . Axe: here Lighthouse is an integrated tool in Chrome DevTools that offers accessibility audits and then gives an accessibility score, among other performance checks. You can learn more about this tool Lighthouse: here. Performing manual testing operations on your application is essential for catching nuanced issues that automated tools might miss. Some manual testing techniques include: Manual Accessibility Testing: Navigate through your application using only the keyboard to ensure all interactive elements are reachable and usable. Keyboard Testing: Use screen reader software (e.g., VoiceOver, JAWS, NVDA) to experience your application as users with visual impairments would. Verify that information is conveyed accurately and interactions are meaningful. Screen Reader Testing: Adjust your browser settings to simulate low vision conditions and test whether your application remains usable. Zoom and High Contrast Testing: Focus Management Testing: Verify that focus indicators (such as outlines) are present and properly styled for keyboard users. Ensure that focus is not trapped in certain elements and that it moves in a logical order. Semantic HTML and ARIA Testing: Review your and verify that you're using appropriate semantic elements (e.g., headings, lists) to convey content hierarchy. HTML structure Check your usage of ARIA attributes to provide additional context to assistive technologies. Real User Testing: Engage users with disabilities in usability testing. Essentially this is also a way to uncover real-world accessibility challenges that automated tools might not detect. Gather feedback on pain points and areas that might need improvements. Conclusion Including web accessibility while building React applications is not an optional consideration; it's an essential responsibility that we all share as web developers. In a world that thrives on digital interactions, ensuring equal access to information, services, and experiences for every individual is paramount. By adopting a proactive approach to accessibility, we can bridge the gap between technology and inclusion, creating a more compassionate and universally usable online environment.

Enhancing Web Accessibility in React Applications

The troubles of the future

Anne of Avonlea by L. M. Montgomery, is part of the HackerNoon Books Series. You can jump to any chapter in this book here . A Golden Picnic Chapter XIII - A Golden Picnic Anne, on her way to Orchard Slope, met Diana, bound for Green Gables, just where the mossy old log bridge spanned the brook below the Haunted Wood, and they sat down by the margin of the Dryad’s Bubble, where tiny ferns were unrolling like curly-headed green pixy folk wakening up from a nap. “I was just on my way over to invite you to help me celebrate my birthday on Saturday,” said Anne. “Your birthday? But your birthday was in March!” “That wasn’t my fault,” laughed Anne. “If my parents had consulted me it would never have happened then. I should have chosen to be born in spring, of course. It must be delightful to come into the world with the mayflowers and violets. You would always feel that you were their foster sister. But since I didn’t, the next best thing is to celebrate my birthday in the spring. Priscilla is coming over Saturday and Jane will be home. We’ll all four start off to the woods and spend a golden day making the acquaintance of the spring. We none of us really know her yet, but we’ll meet her back there as we never can anywhere else. I want to explore all those fields and lonely places anyhow. I have a conviction that there are scores of beautiful nooks there that have never really been seen although they may have been looked at. We’ll make friends with wind and sky and sun, and bring home the spring in our hearts.” “It sounds awfully nice,” said Diana, with some inward distrust of Anne’s magic of words. “But won’t it be very damp in some places yet?” “Oh, we’ll wear rubbers,” was Anne’s concession to practicalities. “And I want you to come over early Saturday morning and help me prepare lunch. I’m going to have the daintiest things possible . . . things that will match the spring, you understand . . . little jelly tarts and lady fingers, and drop cookies frosted with pink and yellow icing, and buttercup cake. And we must have sandwiches too, though they’re not very poetical.” Saturday proved an ideal day for a picnic . . . a day of breeze and blue, warm, sunny, with a little rollicking wind blowing across meadow and orchard. Over every sunlit upland and field was a delicate, flower-starred green. Mr. Harrison, harrowing at the back of his farm and feeling some of the spring witch-work even in his sober, middle-aged blood, saw four girls, basket laden, tripping across the end of his field where it joined a fringing woodland of birch and fir. Their blithe voices and laughter echoed down to him. “It’s so easy to be happy on a day like this, isn’t it?” Anne was saying, with true Anneish philosophy. “Let’s try to make this a really golden day, girls, a day to which we can always look back with delight. We’re to seek for beauty and refuse to see anything else. ‘Begone, dull care!’ Jane, you are thinking of something that went wrong in school yesterday.” “How do you know?” gasped Jane, amazed. “Oh, I know the expression . . . I’ve felt it often enough on my own face. But put it out of your mind, there’s a dear. It will keep till Monday . . . or if it doesn’t so much the better. Oh, girls, girls, see that patch of violets! There’s something for memory’s picture gallery. When I’m eighty years old . . . if I ever am . . . I shall shut my eyes and see those violets just as I see them now. That’s the first good gift our day has given us.” “If a kiss could be seen I think it would look like a violet,” said Priscilla. Anne glowed. “I’m so glad you spoke that thought, Priscilla, instead of just thinking it and keeping it to yourself. This world would be a much more interesting place . . . although it is very interesting anyhow . . . if people spoke out their real thoughts.” “It would be too hot to hold some folks,” quoted Jane sagely. “I suppose it might be, but that would be their own faults for thinking nasty things. Anyhow, we can tell all our thoughts today because we are going to have nothing but beautiful thoughts. Everybody can say just what comes into her head. That is conversation. Here’s a little path I never saw before. Let’s explore it.” The path was a winding one, so narrow that the girls walked in single file and even then the fir boughs brushed their faces. Under the firs were velvety cushions of moss, and further on, where the trees were smaller and fewer, the ground was rich in a variety of green growing things. “What a lot of elephant’s ears,” exclaimed Diana. “I’m going to pick a big bunch, they’re so pretty.” “How did such graceful feathery things ever come to have such a dreadful name?” asked Priscilla. “Because the person who first named them either had no imagination at all or else far too much,” said Anne, “Oh, girls, look at that!” “That” was a shallow woodland pool in the center of a little open glade where the path ended. Later on in the season it would be dried up and its place filled with a rank growth of ferns; but now it was a glimmering placid sheet, round as a saucer and clear as crystal. A ring of slender young birches encircled it and little ferns fringed its margin. “How sweet!” said Jane. “Let us dance around it like wood-nymphs,” cried Anne, dropping her basket and extending her hands. But the dance was not a success for the ground was boggy and Jane’s rubbers came off. “You can’t be a wood-nymph if you have to wear rubbers,” was her decision. “Well, we must name this place before we leave it,” said Anne, yielding to the indisputable logic of facts. “Everybody suggest a name and we’ll draw lots. Diana?” “Birch Pool,” suggested Diana promptly. “Crystal Lake,” said Jane. Anne, standing behind them, implored Priscilla with her eyes not to perpetrate another such name and Priscilla rose to the occasion with “Glimmer-glass.” Anne’s selection was “The Fairies’ Mirror.” The names were written on strips of birch bark with a pencil Schoolma’am Jane produced from her pocket, and placed in Anne’s hat. Then Priscilla shut her eyes and drew one. “Crystal Lake,” read Jane triumphantly. Crystal Lake it was, and if Anne thought that chance had played the pool a shabby trick she did not say so. Pushing through the undergrowth beyond, the girls came out to the young green seclusion of Mr. Silas Sloane’s back pasture. Across it they found the entrance to a lane striking up through the woods and voted to explore it also. It rewarded their quest with a succession of pretty surprises. First, skirting Mr. Sloane’s pasture, came an archway of wild cherry trees all in bloom. The girls swung their hats on their arms and wreathed their hair with the creamy, fluffy blossoms. Then the lane turned at right angles and plunged into a spruce wood so thick and dark that they walked in a gloom as of twilight, with not a glimpse of sky or sunlight to be seen. “This is where the bad wood elves dwell,” whispered Anne. “They are impish and malicious but they can’t harm us, because they are not allowed to do evil in the spring. There was one peeping at us around that old twisted fir; and didn’t you see a group of them on that big freckly toadstool we just passed? The good fairies always dwell in the sunshiny places.” “I wish there really were fairies,” said Jane. “Wouldn’t it be nice to have three wishes granted you . . . or even only one? What would you wish for, girls, if you could have a wish granted? I’d wish to be rich and beautiful and clever.” “I’d wish to be tall and slender,” said Diana. “I would wish to be famous,” said Priscilla. Anne thought of her hair and then dismissed the thought as unworthy. “I’d wish it might be spring all the time and in everybody’s heart and all our lives,” she said. “But that,” said Priscilla, “would be just wishing this world were like heaven.” “Only like a part of heaven. In the other parts there would be summer and autumn . . . yes, and a bit of winter, too. I think I want glittering snowy fields and white frosts in heaven sometimes. Don’t you, Jane?” “I . . . I don’t know,” said Jane uncomfortably. Jane was a good girl, a member of the church, who tried conscientiously to live up to her profession and believed everything she had been taught. But she never thought about heaven any more than she could help, for all that. “Minnie May asked me the other day if we would wear our best dresses every day in heaven,” laughed Diana. “And didn’t you tell her we would?” asked Anne. “Mercy, no! I told her we wouldn’t be thinking of dresses at all there.” “Oh, I think we will . . . a little,” said Anne earnestly. “There’ll be plenty of time in all eternity for it without neglecting more important things. I believe we’ll all wear beautiful dresses . . . or I suppose raiment would be a more suitable way of speaking. I shall want to wear pink for a few centuries at first . . . it would take me that long to get tired of it, I feel sure. I do love pink so and I can never wear it in this world.” Past the spruces the lane dipped down into a sunny little open where a log bridge spanned a brook; and then came the glory of a sunlit beechwood where the air was like transparent golden wine, and the leaves fresh and green, and the wood floor a mosaic of tremulous sunshine. Then more wild cherries, and a little valley of lissome firs, and then a hill so steep that the girls lost their breath climbing it; but when they reached the top and came out into the open the prettiest surprise of all awaited them. Beyond were the “back fields” of the farms that ran out to the upper Carmody road. Just before them, hemmed in by beeches and firs but open to the south, was a little corner and in it a garden . . . or what had once been a garden. A tumbledown stone dyke, overgrown with mosses and grass, surrounded it. Along the eastern side ran a row of garden cherry trees, white as a snowdrift. There were traces of old paths still and a double line of rosebushes through the middle; but all the rest of the space was a sheet of yellow and white narcissi, in their airiest, most lavish, wind-swayed bloom above the lush green grasses. “Oh, how perfectly lovely!” three of the girls cried. Anne only gazed in eloquent silence. “How in the world does it happen that there ever was a garden back here?” said Priscilla in amazement. “It must be Hester Gray’s garden,” said Diana. “I’ve heard mother speak of it but I never saw it before, and I wouldn’t have supposed that it could be in existence still. You’ve heard the story, Anne?” “No, but the name seems familiar to me.” “Oh, you’ve seen it in the graveyard. She is buried down there in the poplar corner. You know the little brown stone with the opening gates carved on it and ‘Sacred to the memory of Hester Gray, aged twenty-two.’ Jordan Gray is buried right beside her but there’s no stone to him. It’s a wonder Marilla never told you about it, Anne. To be sure, it happened thirty years ago and everybody has forgotten.” “Well, if there’s a story we must have it,” said Anne. “Let’s sit right down here among the narcissi and Diana will tell it. Why, girls, there are hundreds of them . . . they’ve spread over everything. It looks as if the garden were carpeted with moonshine and sunshine combined. This is a discovery worth making. To think that I’ve lived within a mile of this place for six years and have never seen it before! Now, Diana.” “Long ago,” began Diana, “this farm belonged to old Mr. David Gray. He didn’t live on it . . . he lived where Silas Sloane lives now. He had one son, Jordan, and he went up to Boston one winter to work and while he was there he fell in love with a girl named Hester Murray. She was working in a store and she hated it. She’d been brought up in the country and she always wanted to get back. When Jordan asked her to marry him she said she would if he’d take her away to some quiet spot where she’d see nothing but fields and trees. So he brought her to Avonlea. Mrs. Lynde said he was taking a fearful risk in marrying a Yankee, and it’s certain that Hester was very delicate and a very poor housekeeper; but mother says she was very pretty and sweet and Jordan just worshipped the ground she walked on. Well, Mr. Gray gave Jordan this farm and he built a little house back here and Jordan and Hester lived in it for four years. She never went out much and hardly anybody went to see her except mother and Mrs. Lynde. Jordan made her this garden and she was crazy about it and spent most of her time in it. She wasn’t much of a housekeeper but she had a knack with flowers. And then she got sick. Mother says she thinks she was in consumption before she ever came here. She never really laid up but just grew weaker and weaker all the time. Jordan wouldn’t have anybody to wait on her. He did it all himself and mother says he was as tender and gentle as a woman. Every day he’d wrap her in a shawl and carry her out to the garden and she’d lie there on a bench quite happy. They say she used to make Jordan kneel down by her every night and morning and pray with her that she might die out in the garden when the time came. And her prayer was answered. One day Jordan carried her out to the bench and then he picked all the roses that were out and heaped them over her; and she just smiled up at him . . . and closed her eyes . . . and that,” concluded Diana softly, “was the end.” “Oh, what a dear story,” sighed Anne, wiping away her tears. “What became of Jordan?” asked Priscilla. “He sold the farm after Hester died and went back to Boston. Mr. Jabez Sloane bought the farm and hauled the little house out to the road. Jordan died about ten years after and he was brought home and buried beside Hester.” “I can’t understand how she could have wanted to live back here, away from everything,” said Jane. “Oh, I can easily understand that,” said Anne thoughtfully. “I wouldn’t want it myself for a steady thing, because, although I love the fields and woods, I love people too. But I can understand it in Hester. She was tired to death of the noise of the big city and the crowds of people always coming and going and caring nothing for her. She just wanted to escape from it all to some still, green, friendly place where she could rest. And she got just what she wanted, which is something very few people do, I believe. She had four beautiful years before she died. . . four years of perfect happiness, so I think she was to be envied more than pitied. And then to shut your eyes and fall asleep among roses, with the one you loved best on earth smiling down at you . . . oh, I think it was beautiful!” “She set out those cherry trees over there,” said Diana. “She told mother she’d never live to eat their fruit, but she wanted to think that something she had planted would go on living and helping to make the world beautiful after she was dead.” “I’m so glad we came this way,” said Anne, the shining-eyed. “This is my adopted birthday, you know, and this garden and its story is the birthday gift it has given me. Did your mother ever tell you what Hester Gray looked like, Diana?” “No . . . only just that she was pretty.” “I’m rather glad of that, because I can imagine what she looked like, without being hampered by facts. I think she was very slight and small, with softly curling dark hair and big, sweet, timid brown eyes, and a little wistful, pale face.” The girls left their baskets in Hester’s garden and spent the rest of the afternoon rambling in the woods and fields surrounding it, discovering many pretty nooks and lanes. When they got hungry they had lunch in the prettiest spot of all . . . on the steep bank of a gurgling brook where white birches shot up out of long feathery grasses. The girls sat down by the roots and did full justice to Anne’s dainties, even the unpoetical sandwiches being greatly appreciated by hearty, unspoiled appetites sharpened by all the fresh air and exercise they had enjoyed. Anne had brought glasses and lemonade for her guests, but for her own part drank cold brook water from a cup fashioned out of birch bark. The cup leaked, and the water tasted of earth, as brook water is apt to do in spring; but Anne thought it more appropriate to the occasion than lemonade. “Look do you see that poem?” she said suddenly, pointing. “Where?” Jane and Diana stared, as if expecting to see Runic rhymes on the birch trees. “There . . . down in the brook . . . that old green, mossy log with the water flowing over it in those smooth ripples that look as if they’d been combed, and that single shaft of sunshine falling right athwart it, far down into the pool. Oh, it’s the most beautiful poem I ever saw.” “I should rather call it a picture,” said Jane. “A poem is lines and verses.” “Oh dear me, no.” Anne shook her head with its fluffy wild cherry coronal positively. “The lines and verses are only the outward garments of the poem and are no more really it than your ruffles and flounces are you, Jane. The real poem is the soul within them . . . and that beautiful bit is the soul of an unwritten poem. It is not every day one sees a soul . . . even of a poem.” “I wonder what a soul . . . a person’s soul . . . would look like,” said Priscilla dreamily. “Like that, I should think,” answered Anne, pointing to a radiance of sifted sunlight streaming through a birch tree. “Only with shape and features of course. I like to fancy souls as being made of light. And some are all shot through with rosy stains and quivers . . . and some have a soft glitter like moonlight on the sea . . . and some are pale and transparent like mist at dawn.” “I read somewhere once that souls were like flowers,” said Priscilla. “Then your soul is a golden narcissus,” said Anne, “and Diana’s is like a red, red rose. Jane’s is an apple blossom, pink and wholesome and sweet.” “And your own is a white violet, with purple streaks in its heart,” finished Priscilla. Jane whispered to Diana that she really could not understand what they were talking about. Could she? The girls went home by the light of a calm golden sunset, their baskets filled with narcissus blossoms from Hester’s garden, some of which Anne carried to the cemetery next day and laid upon Hester’s grave. Minstrel robins were whistling in the firs and the frogs were singing in the marshes. All the basins among the hills were brimmed with topaz and emerald light. “Well, we have had a lovely time after all,” said Diana, as if she had hardly expected to have it when she set out. “It has been a truly golden day,” said Priscilla. “I’m really awfully fond of the woods myself,” said Jane. Anne said nothing. She was looking afar into the western sky and thinking of little Hester Gray. About HackerNoon Book Series: We bring you the most important technical, scientific, and insightful public domain books. This book is part of the public domain. L. M. Montgomery (2022). Anne of Avonlea. Urbana, Illinois: Project Gutenberg. Retrieved https://www.gutenberg.org/cache/epub/47/pg47-images.html This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org , located at https://www.gutenberg.org/policy/license.html. The lead image was generated using HackerNoon's Stable Disffusion Al Image Generator feature, via the prompt "Anne of Avonlea".

A Golden Picnic

I recently sat down with the CEO of BeInCrypto to discuss the rise of crypto journalism and the recent spate of anti-crypto narratives circulating in mainstream media. You can also read Forbes’ story on . Alena Afanaseva How BeInCrypto Took On The Sector’s Macho Image - And Won Does BeInCrypto Have a Physical Office? Thanks to the culture of the Web3 industry, BE[IN]CRYPTO proudly adopts a remote-first and asynchronous work culture. As a result, we have successfully built one of the world's leading crypto media companies, covering 65 markets in 14 languages without needing a physical office or headquarters. Our team, comprising over 260 members across 65 markets, operates entirely remotely. I am based in Argentina, there are opportunities for in-person meetings, and we do occasionally gather with team members at BE[IN]CRYPTO. While I still believe in the importance of face-to-face interactions, I recognize that they are not as crucial as once. Moreover, as a leader committed to sustainability and mindful of our carbon footprint, we embrace a remote-first culture to achieve a net-zero impact. What's your route to blockchain and Web 3? Is crypto journalism a promising industry? I take great pride that BE[IN]CRYPTO's editorial arm has achieved self-sustainability, making us one of the rare news media organizations within the crypto industry and globally capable of building an editorial model that generates sufficient cash flow to maintain profitability. This accomplishment is due primarily to our remote-first approach, which has significantly reduced our financial requirements. However, we are not resting on our laurels. We recognize the strength of our community and readership base and are actively working towards diversifying our portfolio of services. With the emergence of technologies like AI and the changing landscape of content consumption, we are committed to reimagining the future and meeting the evolving demands of our readers. Regardless of any future changes, I firmly believe that the need for a trustworthy news organization with an unbiased narrative will remain as crucial in the future as it is today. Thanks to this strategy, we operate in 14 languages and cover 65 markets while keeping our operational costs lower than our competitors through our remote-first culture. Being financially independent allows us to remain fully transparent and unbiased, enabling us to explore different growth directions, such as developing our education portal, dedicated HR product for talent management, and other services that build our financial strength. As we look towards the future of BE[IN]CRYPTO, our focus remains on diversifying our revenue streams by developing new products that address the needs of industry participants. By continually innovating and providing value for our readers and partners, we are confident that we will maintain our position as a leading media organization in the crypto industry. One such new offering we are developing is an advanced learning portal, providing users with a comprehensive learning experience to develop new skills and gain a competitive edge in the job market through upskilling and reskilling. We aim to bridge the talent gap and fuel adoption by offering a learning and development platform that benefits job seekers and employers. Additionally, we offer managed HR services to top businesses in the industry, facilitating team building, hiring, and talent upskilling. We are confident that our platform will soon become the go-to resource for traditional companies seeking to upskill their existing talent to meet the demands of web3 adoption and development. In addition to vertical diversification, our goal is to become the reference point for everything web3 in all languages. We operate in 14 languages and plan to expand our linguistic capacities to provide even more coverage. BE[IN]CRYPTO offers services in more languages than any other direct competitor, giving us an edge in the industry. How did you set up Beincrypto? In 2018, the idea for BE[IN]CRYPTO was born during the ICO boom. With a decade of experience as a fundamental analyst, I saw the vast potential of blockchain projects and recognized a dire need for reliable information in the industry. Many inexperienced entrepreneurs were attempting to cash in on the burgeoning industry, leading to many scams, misinformation, and shilling around crypto. I had the right background and experience to develop a crypto news media project that could fill this gap. Our mission to simplify the first steps for non-experienced users set us apart from competitors, making the web3 space accessible for everyone. We focused on providing trustworthy and accurate information for all things crypto, aimed specifically at newbies. This approach proved successful, and within six months of launch, we attracted 500K monthly users. We soon realized that while crypto market information was accessible in English, it was scarce in other languages. This prompted us to change our focus from growing within the English-speaking audience to expanding worldwide, offering more opportunities to people from emerging markets. After a year of operation, we formed local teams to provide original content in their language, starting with Brazilian Portuguese, Spanish, Russian, and Turkish. Two years later, we added four more languages, becoming the world's leading international crypto new media. In 2021, we became the first crypto news media to adopt globally recognized trust indicators through the 'Trust Project.' The project honors news sites that comply with its 8 Trust Indicators aiming for news transparency and accuracy. This recognition helped us establish ourselves as a reliable source of information and cemented our commitment to providing only fact-checked information. Today, we publish unique content in 14 languages written by locals who understand the audience's pains. At BE[IN]CRYPTO, we strive to empower your web3 journey, regardless of your background and skills. Our mission is to make the web3 space accessible to everyone, and we will do our best to provide you with all the needed information without misleading or complicating things. While the costs were on us for the first six months after we launched BE[IN]CRYPTO, we continue to expand and provide a platform for all things crypto. What were the early challenges you faced setting up your media company? How did you overcome them? At BE[IN]CRYPTO, we lead by example when it comes to finding top talent. We believe in hiring individuals who share our values of responsibility and self-drive, are passionate about new technologies, and are willing to work hard to build something from scratch. However, finding high-quality talent can be challenging, especially when working with a limited budget and in a new industry like web3. When we started, we faced the same hurdles, but through perseverance, we built the right team of 250+ employees spread across 50+ countries. Managing a 100% international and remote team from different cultures, backgrounds, and time zones in a decentralized setting is an ongoing improvement process. Local teams produce our content, and we have partners in over 160 cities worldwide. These regional teams take timely global information and fine-tune it to their target audience, speaking their language - literally. Building on this success, we are now developing our HR product offering to help similar businesses in the industry build the right capacities in a decentralized and asynchronous manner. Through our expertise in managing a remote team, we aim to help others navigate this new way of working and build their successful decentralized teams. Does crypto journalism have a common philosophy? Bloomberg, Financial Times, and CNBC are increasingly entering the crypto sphere by establishing dedicated news branches. While several successful crypto information portals cater to various needs, we closely follow the work of our colleagues at Decrypt, Cointelegraph, and Coindesk. Despite the intense competition, we take pride in that BE[IN]CRYPTO, the youngest among them, has established itself as one of the top three players in less than four years. At BE[IN]CRYPTO, we differentiate ourselves by simplifying the initial steps for non-experienced users and helping a less mature audience understand the web3 and crypto industry, explore its potential, and embark on their journeys. This sets us apart from competitors who primarily target the needs of experienced "crypto geeks," making the space exclusive and inaccessible to a broader audience. Our mission is to empower individuals on their web3 journey, regardless of their background and skills. We firmly believe that anyone can enter the space and are committed to providing all the necessary information without any misleading or complicated jargon. Our goal is to make web3 accessible to everyone by offering learning opportunities and employment possibilities within the industry, allowing individuals to develop the skills and knowledge needed to thrive in the crypto world. Today, we can confidently assert that BE[IN]CRYPTO and most of our direct competitors in the industry take journalism very seriously and adhere to the highest journalistic ethics when collecting and publishing news. However, as the sector is forward-thinking, we must also be agile and creative. It has become common for media outlets to reference tweets from insiders, access community information, and build stories around bite-sized knowledge. This agile approach is driven by the need for financial market players to have firsthand information to make informed decisions. As a result, I believe that crypto media is challenging the status quo and traditional media while maintaining trustworthiness and a solid reputation. Is it reactive to mainstream media pushing a wrong narrative about crypto? Mainstream media has long been influenced by bias shaped by sponsors and external funding. As I mentioned earlier, at BE[IN]CRYPTO, we firmly believe that only financially self-sustainable media organizations can maintain unbiased reporting in today's complex entrepreneurial environment. Traditional media has exhibited bias not only in their coverage of crypto and web3-related news and stories but also in the handling of incidents like the FTX scandal, where media outlets tended to approach the story with a victim-blaming perspective, failing to hold criminals accountable and using language that softens the severity of the situation. The New York Times, for instance, was among those guilty of this approach. In other words, it is exhibits bias against crypto. I firmly believe that any media organization should recognize its power and act with goodwill, serving as an open platform for education and enlightenment without bias. It may sound overly optimistic, and perhaps it is, but this is how we operate within our organization. Where do you see it heading? As I mentioned earlier, the trustworthiness of a news media organization cannot be achieved when it heavily depends on external factors. At BE[IN]CRYPTO, we take pride in being an independent news media organization with a positive cash flow. We have chosen a different path in an industry where many players receive heavy subsidies, such as Coindesk, owned by Digital Currency Group, and Binance, investing in news media organizations. Binance, for example, made a $200 million investment in Forbes shortly after suing the publisher for defamation and even secured seats on the media organization's board. Instead of relying on external financial support, we believe in sustainable and steady growth, reinvesting our profits into developing markets. The question of how a news media organization can remain unbiased when it heavily relies on external investments for survival is not unique to crypto media but applicable to the entire news business and media industry. While financial support and funding may offer a competitive advantage to some businesses, our focus on sustainable growth allows us to maintain independence and make decisions that prioritize the needs of our readers and the markets we serve. This strategy also enables us to attract top talent, enhance our editorial capabilities, and expand our user base in alignment with our values and mission. We have established several revenue streams at BE[IN]CRYPTO contributing to our positive cash flow. Our primary source of revenue comes from advertisers and affiliate partners, reflecting the value of our platform as a leading source of information and insights for the crypto industry. In addition to advertising, we have experienced a surge in demand for our expert HR product, which assists emerging Web3 businesses in building their teams, streamlining hiring processes, and establishing remote work foundations. Our job portal, featuring opportunities from top employers in the industry, has also seen substantial growth, attracting over 20,000 job seekers every week. Our newly developed education portal and consulting business have started generating revenue after years of steady investment and product development. These areas hold significant potential for further growth and expansion in the coming years. Looking ahead to 2023, our goal is to diversify our revenue streams through new product development that caters to the needs of industry participants. We are confident in maintaining our position as a leading media organization in the crypto industry by continuously innovating and introducing offerings that add value for our readers and partners.

Stories, News & Mentions about WEGROUP

Wegroup’s Stories on HackerNoon

Wegroup's News & Mentions