UpCodes is unveiling a new AI-based tool Copilot that will further streamline the world of writing code. Copilot, based on GPT-4, serves as a research assistant by responding to complex code questions and providing explanations with links to pertinent code sections.
UpCodes’ primary goal prior to the release of Copilot was to create its own database of codes, digitize regulations that were frequently only available in printed reference materials, and make them simple to examine. It contains 160,000 local modifications in addition to five million code parts. Since codes are continually changing, UpCodes releases an average of 7,000 changes per month.
Copilot contains a searchable database and additional tools, like its code check feature, that are intended to make code compliance simpler, but because of the complicated regulations, these are also difficult to use at first. Copilot wants to make the process of finding new codes incredibly simple. Copilot responds to these inquiries and assists users by listing the code source from which it derives the same so they can examine the actual code for themselves.
For more than seven years, UpCodes has helped both professionals in the industry and amateurs better grasp the complicated world of building codes. On its website, you may explore a library of laws from every state and take advantage of tools like a “spell check” that highlights coding mistakes.
Additionally, Upcodes disclosed that it has closed a $3.5 million Series A round of funding with the intention of developing Copilot and expanding the platform’s AI-based features. UpCodes has now raised $7.6 million in total, including prior funding and a Pre-Series A that was announced in March 2021.
Building Ventures, a VC firm specializing in construction and real estate tech, took the lead in the most recent round. CapitalX, Bragiel Bros., and the co-founders of PlanGrid are other participants.
University of Hong Kong has introduced Text2NeRF, a text-driven 3D scene synthesis system that combines the Neural Radiance Field (NeRF) and the best characteristics of a trained text-to-image diffusion model.
Researchers picked NeRF as the 3D representation because of its superiority in modeling fine-grained and lifelike characteristics in a variety of circumstances, which may significantly reduce the artifacts generated by a triangular mesh. They replaced older methods like DreamFusion, which used semantic priors to govern the 3D creation, with finer-grained picture priors inferred from the diffusion model.
Because of this, Text2NeRF can generate realistic texture and delicate geometric shapes in 3D scenes. A pre-trained text-to-image diffusion model is used as the image-level prior, and they constrain the NeRF optimization from scratch without the requirement for additional 3D supervision or multiview training data.
Priors for depth and content are used to optimize the NeRF representation’s parameters. To be more explicit, they build a text-related picture as the content prior using a diffusion model and a monocular depth estimation approach to offer the geometric prior of the constructed scene. In order to guarantee consistency across numerous viewpoints, they also recommend a progressive inpainting and updating technique (PIU) for the unique view synthesis of the 3D scene.
Text2NeRF developed a variety of 3D settings, including artistic, indoor, and outdoor scenes, due to the method’s universality. Text2NeRF can also create 360-degree views and is not limited by the view range. Their Text2NeRF performs qualitatively and statistically better than the preceding approaches, according to numerous tests.
‘Odisha for Artificial Intelligence‘ and ‘Artificial Intelligence for Youth‘ programs were introduced by Odisha Chief Minister Naveen Patnaik on Monday in the state capital. International technology company Intel has been enlisted by the State Government for the project. The programme would be put into action in Bhubaneswar, Puri, and Cuttack during the initial phase.
According to Tusharkanti Behera, the state’s minister of electronics and information technology, this Artificial Intelligence project will elevate Odisha to the top tier of Indian states. On the official government website, Intel offers a free 4-hour AI training called Odisha for AI. In addition, it will be available to everyone in Odisha. AI for Youth is for students below 18 in all 2000 schools that fall under the 5-T initiative and Odisha Adarsha Vidyalayas. It will be available to everyone for free in Bhubaneswar, Cuttack, and Puri.
Speaking at the event, the Chief Minister said artificial intelligence has the power to drastically change the way we live and advance society. He said that one of the core elements of the government’s 5-T initiative, technology-driven transformation, has been the focus of his government. He pledged that the effort will increase the general public’s digital literacy and familiarize them with the most cutting-edge technology of the next generation. He continued, “It will also establish an environment that supports research, innovation, and application across sectors.
He praised the State Electronics and IT Department and Intel India for their cooperation and urged all government agencies to benefit from these initiatives and train their employees so they can lead the revolution in artificial intelligence. He expressed optimism that the effort would use artificial intelligence to improve society and the lives of citizens while also fostering economic growth and transforming governance.
In order to unleash the boundless potential of youth and create a future where artificial intelligence is a tool for the empowerment of citizens and inclusive progress of the state, the CM urged everyone to embrace this new chapter. Moreover, Tusharkanti Behera emphasized how technology is reshaping Odisha in a number of different areas.
According to sources, TikTok is developing a chatbot named Tako that uses artificial intelligence to recommend videos based on questions users ask it.
The chatbot might “radically change search and navigation” in the app if TikTok decides to make it broadly available, according to Daniel Buchuk of Watchful Technologies, a company that foresees these kinds of forthcoming app modifications for Fortune 500 firms.
Tako can be seen seated above the TikTok profile icon to the right of a video in screenshots from the test that Buchuk shared. When you tap it, a chat box appears where the bot looks to be able to respond to a variety of questions. What AI model TikTok is utilizing to power Tako remains unknown.
To assist a user in starting a discussion with the bot, Tako will show suggested prompts. Buchuk claims that “if I’m watching food videos and ask for a recipe, I’ll get related TikTok videos for the recipe, or if I ask for good art exhibitions in Paris, it’ll show videos alongside a list of suggestions in the answer.”
1/ We're in the early stages of exploring chatbot tools with a limited test of Tako with select users in the Philippines. Tako is an AI-powered tool to help with search and discovery on TikTok.
Zachary Kizer, a representative for TikTok, described the chatbot as “a limited experiment” and informed that users in North America or Europe aren’t yet allowed to use it. The business stated in a tweet that the test is exclusively taking place in the Philippines.
NVIDIA unveiled the NVIDIA Avatar Cloud Engine (ACE) for Games at COMPUTEX 2023, which will usher in the era of non-player characters (NPCs). With the help of AI-powered natural language exchanges, NVIDIA ACE for Games, a bespoke AI model foundry service, intends to alter games by giving NPCs intelligence.
NVIDIA ACE for Games enables middleware, tool, and game developers to create and integrate specialized voice, communication, and animation AI models.
At Computex 2023 in Taipei, Nvidia CEO Jensen Huang recently showed the world a peek of what it would be like when gaming and AI come together with a graphically stunning rendering of a cyberpunk ramen shop where you can actually interact with the owner.
Rather than selecting conversation options, it imagines that you may hold down a button, simply speak with your own voice, and receive a response from a video game character. It is what Nvidia refers to as a “peek at the future of games.”
In order to demonstrate how developers will soon be able to use NVIDIA ACE for Games to generate NPCs, NVIDIA teamed together with Convai, an NVIDIA Inception business. Convai integrated ACE modules into their end-to-end real-time avatar platform. Convai is a company that specializes in creating cutting-edge conversational AI for virtual game worlds.
According to John Spitzer, vice president of developer and performance technology at NVIDIA, “Generative AI has the potential to revolutionize the interaction that players can have with game characters and dramatically increase immersion in games.”
Technology Innovation Institute (TII) created Falcon-40B, a potent decoder-only model that was trained on a huge amount of data made up of 1,000B tokens from RefinedWeb and curated corpora. The TII Falcon LLM Licence makes this model available publically. This open-source model, on the OpenLLM Leaderboard, outperforms models like LLaMA, StableLM, RedPajama, and MPT.
The Falcon-40B’s inference-optimized design is one of its noteworthy features. It features multi-query, which Shazeer et al first detailed in 2019, and FlashAttention, which Dao et al first introduced in 2022. The model performs exceptionally well and is extremely effective when doing inference tasks as a result of these architectural improvements.
Another one is Falcon 7-B. A very sophisticated causal decoder-only model called Falcon-7B was created by the Technology Innovation Institute (TII). It has been trained on a sizable dataset of 1,500B tokens obtained from RefinedWeb, further augmented with curated corpora, and has an outstanding parameter count of 7B. The TII Falcon LLM Licence is used to make this model available.
Falcon-7B’s superior performance in comparison to other comparable open-source models like MPT-7B, StableLM, and RedPajama is one of the main justifications for picking it. As seen on the OpenLLM Leaderboard, its better skills are a result of extensive training on the enriched RefinedWeb dataset.
An architecture that is specifically designed for inference tasks is included in Falcon-7B. The model gains from the incorporation of multi-query, introduced by Shazeer et al in 2019, and FlashAttention by Dao et al. Same as Falcon-40B, the model is more effective and efficient while performing inference operations as a result of these architectural improvements.
This week, JPMorgan Chase plans to layoff around 500 workers across the bank, according to a corporate representative, as reported by CNN Business.
According to the article, which cited JPMorgan, the layoffs will occur throughout the whole corporation but will mostly target the verticals of operations and technology. As per the JPMorgan representative, the company continues to hire while reviewing its business and client needs on a regular basis.
The bank employs close to 300,000 employees and has about 13,000 unfilled positions. In the fields of commercial banking, investment banking, processing financial transactions, and asset management, JPMorgan is a world leader.
Just one day after JPMorgan told roughly 1,000 First Republic Bank workers that their positions would be eliminated, the news of the most recent 500 job cutbacks broke. After the US government seized the regional bank with headquarters in San Francisco, JPMorgan bought First Republic Bank earlier this month.
First Republic Bank, a US-based institution, was shut down by local regulators on May 1 and, in an effort to safeguard depositors, entered into a deal with JPMorgan Chase Bank to buy and assume all of the ailing bank’s deposits and assets.
When raw data is collected, processed, and analyzed using various machine learning models and AI-powered algorithms, valuable insights emerge. These insights help organizations improve their understanding of business, identify threats and opportunities, and bring efficiency to operations. Using data analytics, organizations can derive many types of insights from data. Some insights provide straightforward answers to questions, some insights provide confirmation for a hypothesis, and some insights predict the future course of business events. And then, there are insights that stimulate action and accelerate decisions. Such insights are called actionable insights.
What are Actionable Insights?
Actionable insights are insights that deliver immense value, provide a clear direction, and drive instant actions towards achieving business objectives. These insights are the most valuable and provide maximum returns on data analytics investments.
Actionable insights are distilled from raw data in the following way.
Data: Raw data is collected from multiple sources such as information databases, spreadsheets, plans, social media comments, call logs, and so on. This data can be structured (forms and tables) or unstructured (log files, comments) and qualitative (observations) or quantitative (numbers or units).
Information: Raw data, when processed, categorized, and organized, is converted into information. This information can be presented in the form of reports, dashboards, charts, and other visualizations.
Insight: Information is analyzed to identify patterns, summarize findings, and draw conclusions in the form of insights. Insights help to understand business events, answer questions, compare details, spot similarities and differences, group similar items, and track trends.
Actionable Insight: Among the various insights that emerge from data, actionable insights are crucial. These insights encourage informed actions and decisions based on data-driven recommendations. The control to take action or not rests with the user.
For example, the photo gallery app accesses and collects different media files (raw data) stored in various folders of the phone’s internal memory and extended memory in one place. It classifies this data into categories by format, size, file type, timestamp, or tags and presents them visually in the form of tiles, bar charts, or badges (information). It indicates that some files are old, unclear, very large in size, or duplicate (insight). It offers the option to delete duplicate files, fix unclear or hazy images, and indicates how much space will be saved or gained back after deleting files with large sizes (actionable insight).
Benefits of Actionable Insights
Actionable insights help make sense of business events and take action quickly and confidently. In a fast-changing market, identifying threats and converting opportunities early with actionable insights gives organizations a significant competitive edge.
Enhance Decision Making
Actionable insights give business users the much-needed support to enhance their decision-making. When insights are aligned with an organization’s business objectives and KPIs, they become more actionable and influence better decision-making. For example, knowing every increase or decrease in social media views might be a vanity metric if the marketer’s actual goal is to track click-throughs that lead to the company’s e-commerce website and convert it into sales. Actionable insights provide decision intelligence on what matters most to decision-makers based on their role, work, or goal.
Understand Context Better
Actionable insights go beyond answers to provide context for changing business metrics. For example, the monthly report shows a surge in sales by 20% over the previous month. It may look good now, but it is missing a context. The surge was due to an increase in the price of goods, while the number of transactions actually went down. Such a surge might not help in the long run. This actionable insight alerts the management that the price rise is driving away customers, and they can take measures to win back customers. Actionable insights explain not only what happened but also why it happened and how to proceed ahead.
Gain Clarity
Processing data and extracting insights can be managed with automated data analytics, machine learning models, and AI-powered algorithms. However, presenting the insights clearly in an understandable manner makes them more actionable and relevant. Actionable insights, when presented as interactive charts with multiple visualizations, give users control over how they want to see it. Bite-sized audio-visual data stories make actionable insights not only easily consumable but also engaging.
Get Specific Recommendations
Actionable insights cut through the noise of lengthy reports, stacks of dashboards, and information overload to provide a clear direction. When users receive specific, actionable insights, there are higher chances that they will act on them soon. For example, actionable insight on the increase in demand from a new segment of customers delivered in the form of automated business headlines helps customer success managers devise targeted offerings and create better relationships.
Save Significant Time and Efforts
When users get access to relevant, focused, and personalized insights, they can understand business situations accurately and make quick decisions. With actionable insights, users no longer have to go through volumes of information and unnecessary vanity insights. With modern data analytics platforms’ intuitive search interfaces like MachEye, users can access data easily and get insights in an automated way without investing time and effort in learning complex techniques or depending on experts.
Transform Raw Data into Actionable Insights with Machine Learning
Machine Learning helps in automating time-consuming data processing tasks, unearthing actionable insights, and predicting trends from enterprise data. With the right combination of machine learning models and modern data stacks, organizations can transform raw data into actionable insights efficiently.
Define Clear Objectives
It is important to state the business problems that users are trying to solve with data analytics. Set clear expectations on the outcomes and insights users want to receive from the analysis. Identify and track the KPIs and business metrics that are most relevant for getting actionable insights. Ensure that they are measurable and aligned with the business objectives.
Identify Data Sources and Connect to Data
The data to be analyzed can be present in different data sources and formats, generated and maintained across different departments within an organization. Identify the data that are needed for analysis and integrate it for a comprehensive view. Using a modern analytics platform, data from multiple sources can be connected directly without the trouble of duplicating or frequently loading it. This also ensures that current and relevant data is available to generate the latest and most accurate insights.
Prepare and Model Data for Exploration
Collecting and preparing data accounts for the majority of time and effort in the process of distilling actionable insights from data. Modern analytics platforms such as MachEye not only automate data preparation and curation but also measures its quality and offer recommendations to fix data observability issues. Machine Learning plays a critical role in modeling data to achieve the desired insights. Select and evaluate the right ML models that suit the specific needs of an organization. A modern analytics platform that offers robust models and also supports out-of-the-box or customized models gives organizations greater flexibility and control over their analysis.
Set Benchmarks and Provide Context
Success can be measured differently by different organizations based on their objectives. It can be more store footfalls for some, whereas it could be an increase in staff productivity for others. Set a benchmark for what success criteria look like. Provide a context that explains the significance of numbers or values. Is a 10 % increase good, average, or bad? What parameters should products fulfill to be ranked in the top 5 list? Machine Learning algorithms customized based on such context can yield faster and better results. For example, banks and financial lending institutions use ML algorithms to evaluate a loan applicant’s creditworthiness and send actionable insights to loan officers on whether to approve or reject the application.
Recognize Patterns and Deliver Actionable Insights
Various Machine Learning methods can be employed to recognize patterns, draw analogies, identify anomalies and outliers, trace clusters, and make predictions. Configure these models to address specific insight needs. To make the insights actionable, it is necessary to present them in the right format at the right time so that users can act on them soon. Modern analytics platforms such as MachEye present actionable insights in the form of interactive charts, engaging audio-visual data stories, and automated business headlines. Analytics platforms that learn from their users can deliver personalized and focused insights whenever they happen in data and not just when users search for them.
Conclusion
In conclusion, actionable insights are no longer a luxury restricted to the higher echelons of management. They have become a necessity for every decision-maker across the organization, be it on the frontline, on the shop floor, or in the back office. With modern data analytics platforms powered by AI and ML models and technologies, the extracting, presenting, and consuming of actionable insights has become simplified, personalized, automated, and accessible for everyone.
On Thursday, May 25, forty education ministers from around the world and the UN Educational, Scientific and Cultural Organisation (UNESCO) virtually gathered to go over the advantages and disadvantages of utilizing chatbots in the classroom.
According to UNESCO, less than 10% of schools and universities adhere to official guidelines when employing AI technologies like the chatbot programme ChatGPT. Ministers had the chance to discuss policy philosophies and strategies for the secure and efficient use of AI in education during the meeting.
“Generative AI opens up new horizons and challenges for education, but we urgently need to take action to ensure that new AI technologies are integrated into education on our terms,” said Stefania Giannini, Assistant Director-General for Education at Unesco. “We have a responsibility to put safety, inclusivity, diversity, transparency, and quality first.”
The meeting brought up a number of issues that participants had with chatbots being used in classrooms. One worry is that chatbots may make obvious mistakes. How to incorporate these tools into curriculum, teaching strategies, and tests is another issue.
Ministers also talked about how education systems need to change to accommodate the disruptions that generative AI is already causing. Many ministers emphasized the crucial function of teachers as learning facilitators in this new era. However, according to Unesco, instructors require direction and training to handle these difficulties.
UNICEF will continue to lead the international conversation with partners, politicians, academics, and members of civil society. The organization is also creating frameworks of AI competencies for teachers and students, as well as policy recommendations on the use of generative AI in research and education. These new resources will be unveiled during Digital Learning Week, which will take place from September 4 to 7 at the UNESCO headquarters in Paris.
The NVIDIA DGX supercomputer, powered by NVIDIA GH200 Grace Hopper Superchips and the NVIDIA NVLink Switch System, was unveiled by NVIDIA. It was designed to support the creation of enormous, next-generation models for generative AI language applications, recommender systems, and data analytics workloads.
The NVIDIA DGX GH200’s enormous shared memory area combines 256 GH200 superchips into one GPU using the NVLink connection technology and NVLink Switch System. This offers 144 terabytes of shared memory and 1 exaflop of performance, which is roughly 500 times more memory than the 2020-released NVIDIA DGX.
By integrating an Arm-based NVIDIA Grace CPU and an NVIDIA H100 Tensor Core GPU in the same device and connecting them with NVIDIA NVLink-C2C chip interconnects, GH200 superchips do away with the requirement for a conventional CPU-to-GPU PCIe connection. This offers a 600GB Hopper architecture GPU building block for DGX GH200 supercomputers, reduces interconnect power consumption by more than 5x, and boosts GPU and CPU bandwidth by 7x compared to the most recent PCIe technology.
By offering 48 times more NVLink bandwidth than the previous generation, the DGX GH200 architecture combines the simplicity of programming a single GPU with the power of a powerful AI supercomputer. The DGX GH200 is scheduled to be made available to Google Cloud, Meta, and Microsoft first, allowing them to test its potential for generative AI workloads.
In order to allow cloud service providers and other hyperscalers to further customize the DGX GH200 architecture for their infrastructure, NVIDIA also plans to make the DGX GH200 design available to them as a blueprint. Supercomputers powered by the NVIDIA DGX GH200 are anticipated to become available by the end of the year.
