While Ola and Porter scored zero in fair work standards for gig workers. and bigbasket got six points, none of the 12 platforms evaluated on fair pay, work conditions, contracts, management, and representation principles managed to get a perfect score of 10.

Ola, Uber and Dunzo among worst platforms for gig workers, bigbasket tops fair work norms list

BigBasket is the best place for gig workers, while Ola and Porter are the worst, according to research by Fairwork India

Unfair pay, poor working conditions, and unresponsive management: From BigBasket to Ola, here are the best and worst companies for gig workers

While Dunzo and Uber scored only one point out of 10 on Fairwork’s 2023 ratings, Ola and Porter scored zero points

Ola, Uber, Dunzo, Porter Among Worst Performers On Fairwork’s Rating For Gig Platforms

BigBasket has claimed the top spot, while ride-hailing giants Uber and Ola, along with logistics startup Porter, have been rated as the least favourable platforms for gig workers, as per a study by research firm Fairwork India.

Ola, Uber And Dunzo Among Worst Platforms For Gig Workers: Survey

Of all the digital platforms evaluated by the Fairwork India Team, only BigBasket scored the highest point 6 out of 10 across five principles (fair pay, fair conditions, fair contract, fair management and fair representation) for gig workers.

Ola, Porter, Uber, Dunzo score poorly on fairness ratings by Gig workers; BigBasket highest

The report was structured around the theme of discrimination and alienation that often come up in experiences shared by platform workers

Ola, Porter, Uber & Dunzo worst digital platforms for gig workers, says report

Both Dunzo and Uber have scored one point each. Last year, Urban Company scored seven points and topped the ranking, followed by BigBasket with 6 points. Also, this year, the Fairwork India team has included a ride-hailing platform aggregator that owns ...

Ola and Porter score zero in Fairwork rating for gig workers

A majority of digital platforms in India do not have a minimum wage policy in place for gig workers, pointed out a new report titled Labour Standards in the Platform Economy, by Fairwork India. ADVERTISEMENT

Most Indian platforms are silent on their minimum wage policy for gig workers: Report

Bigbasket, Flipkart, and Urban Company have implemented a minimum wage policy to ensure their workers earn the local minimum wage. The Fairwork India

Only three online companies offer minimum wage: Study

BluSmart, Swiggy, Urban Company, and Zomato scored five points and Zepto four. Ola and Porter scored a duck (zero) on the Fairwork India Ratings 2023, which ranks the companies on the minimum standards of fair work for gig workers.

BigBasket gets highest score, Ola and Porter score zero in fair working conditions for gig workers

Dunzo

dunzo

Product management is an exciting and challenging field. in a marketplace, though, is a completely different world with its own unique ups and downs. While there are best practices around what to do during the scale-up or 1 to 100 phases, I am here to talk about what aspects of the product we should pay attention to during the zero-to-one product management at marketplaces. Product management Throughout my 10+ year career in marketplaces, I was involved in building three businesses. In my previous role at , I worked as a product manager responsible for launching Uber Fleet products globally. While at Dunzo, a delivery app, I led the product launch and expansion team for a ride-sharing product, Dunzo Bike Taxi, in India. I’m currently leading the Platform and Data Engineering Product team at Carwow, an online marketplace for buying and selling cars in London, UK. I’d like to talk about how a product manager can add value during the launch and the zero-to-one phase and share some advice based on my experience. Uber First, what is a marketplace, and how does it function? In short, a marketplace is an intermediary that connects demand (end consumers) and supply (sellers or vendors). Recently, technology has accelerated this type of model - think Uber connecting commuters with taxis or Doordash connecting consumers with restaurants. The marketplace, in turn, keeps a percentage of the transaction and calls it a commission or service charge. So, how does product management fit in? While it may seem like the role of a product manager does not really change with the phase of growth of an organization, a marketplace might demand an entirely different mindset. More akin to a wartime CEO and peacetime CEO, there could be a Launch Product Manager and Growth Product Manager at a marketplace. ( if you are not venture-funded): The success of a newly launched marketplace is determined by two things three - The beauty of marketplaces is how they solve the immediacy of a particular demand. In the Uber example - you want to book a cab immediately rather than days in advance. However, this means keeping demand and supply in sync due to the transient nature of both. Building tools to quickly onboard suppliers and make it easy for customers to sign up is something that products can add plenty of value to. Liquidity - If more people use and like the product, the higher the chance they will recommend it to their friends. This in turn creates more demand for sellers or suppliers, thereby increasing the revenue they earn. This can be done through referrals or via user promotions. Both are areas that require product teams to drive. Network effects (if sustainability is a focus during launch) - If unit economics is a priority, having tools to configure pricing, user promotions, seller incentives, etc., becomes more important during the launch phase. However, typically, these are areas that venture-funded companies focus on later in the journey after finding the product market fit. Unit Economics In short, you need a strong product team to tackle all three parts of the marketplace. How can a Product Manager add value during launch and the zero-to-one phase? However, a lot can go wrong with how product managers handle a launch. Assuming you are the PM, or a small group of PMs, involved in launching a new marketplace, or a new category or geography within an existing marketplace, here are things that you should be doing. - Good enough is excellent, and excellent is overkill when it comes to the design and the UI during launch. Users and sellers will not care what color your logo is as long as they can get the intended outcome from the platform (e.g., moving from point A to B for consumers and earning additional income for suppliers). Keep it simple and use the limited resources elsewhere, like the user experience (UX). Yes, UX and UI are not the same, and I learned this the hard way. While building the Uber Fleet app, we tested a few iterations of the UI with end users (fleet partners), and all they cared about was whether they could download a .csv file from the dashboard. No one talked about the font used or the colour of the button. Do not go overboard with the interface - While the demand-supply equation of a marketplace is very dynamic, there will always be one side to optimize for. Either you have higher demand than supply (more customers than cabs for Uber) or higher supply than demand (more cabs than demand or customers). The marketplace’s job is to balance both sides while trying to grow their profit (or revenue if pre-profit). Product managers must double down and solve this to improve liquidity. There is no sense in optimizing the consumer flow if there are not enough cabs on the road. Focusing on onboarding cabs faster might be a better use of a PM’s time. While launching Dunzo, an on-demand marketplace, we realized this when we had a lot of customer demand but not enough restaurant partners on board, as we had not put enough effort into building the partner onboarding flow. Focus on the right side of the marketplace - This is as important during launch as it is later. Your first customers and partners will be your power users in the future and drive network effects later. Knowing and prioritizing their pain points will lead to a better output than building a fancy supply-demand matching algorithm, for example. Trust and safety were a big challenge at Uber, and one of the ways the product could help was by talking with customers and partners regularly. We even had a WhatsApp group of power users at Dunzo who helped give us feedback and identify loopholes in our user promotions. Don’t keep customer or partner feedback for later - Building internal tools is part of launching a marketplace. These are internal tools that consumers or suppliers sometimes don't even see or are aware of. It’s essential to keep them simple and not build everything at once. An interesting use case I found was building an internal promotions tool to configure discounts. It might be wise to just start with flat percentage discounts instead of trying to develop all types of user discounts at once (e.g., buy 2 get 1 free or order 5 times and get the 6th order free). It is important to build a limited set of polished features instead of creating multiple features with inadequate checks and balances. Quality over quantity - Analytics should be everyone’s job during launch. However, there is no point in tracking or analyzing data that is not driving the marketplace to grow. For example, the number of customers ordering ice cream between midnight and 4 a.m. is not driving many product decisions during launch unless you are an ice cream marketplace with no supply at night. And even then, it goes back to point #2. Don’t track everything - While experimentation is valuable for improving the product, you might not have enough data available during the launch phase. Keep the experiment simple and quick, if possible, and don’t get obsessed with experimenting with everything. Relying on your gut or instincts might be better at this stage; however, it is not recommended with significant changes once the company has found a product market fit. Experiment, learn, and move on A few other things to keep in mind for Product Managers The product roadmap should be based on the direction the business wants to take. For example, if the business strategy is to acquire enterprise suppliers instead of small or independent contractors, the product should reflect that. Stay connected to the business strategy. It is vital to let stakeholders know what’s coming. Often, product changes need to be complemented by changes to other processes that the product team might not be even aware of. For example, customer service FAQs. Keep everyone in the loop. Keeping a close relationship with a few power users and partners will shorten the feedback cycle and help iterate and improve the product faster. Get involved with your customers and partners. It is imperative that at least the leadership team are power users of the marketplace. In the case of a B2B or low-frequency marketplace, it is important to keep testing the user flow through a test account that everyone can access. There are a lot of valuable insights and feedback that can be derived from dog-fooding, especially in the early stages. Encourage dogfooding. Overall, being a product manager during launch is like walking a tightrope over Niagara Falls. It is pretty rewarding when done right, however deadly if you focus on the wrong things. I hope you found these tips valuable and please let me know if you agree or disagree with them.

Zero-to-one Product Management in Marketplaces 

When one seeks to distinguish between Web2 and Web3, a key value that differentiates the two is the concept of ownership. Simply put, what you create is what you own and can monetize. Just as it should be. Nothing less, nothing more. In fact, the day you mint your first NFT is when you’ve made it in Web3 as an owner, thanks to the immutability of the blockchain. If anything, the feeling of being protected is priceless. Speaking of which, this concept of ownership is just as important to smart contracts in respect to accessible functions and permitted state change. Your First Look at Ownership with Smart Contracts So, why does it matter that you ‘own’ the smart contract that you’ve written? Think about it: if you owned a bike, would you hesitate to keep ownership records? Just in case someone stole and used it for nefarious purposes? Of course, not. It’s for the very same reason why you would want to prove your ownership of an NFT or a smart contract. Well, someone might just gain unauthorized access and benefit financially from it, right? Let’s say you work with Dunzo and have written this smart contract where you can set the package price plus delivery charges along with whether it was paid or not by the buyer. In the ideal scenario, you can set the buyer’s name, the total price of the package and delivery to a certain amount and set the setPackageDelivered to true. Once you’re done delivering the package, of course. But, as we all know, nothing ever goes to plan. Now, what if the buyer had access to the contract functions and the ability to reset the packageDelivered value to false or change the packageDeliveryPrice value to a lesser one? Not only can they pay less for the product but get away with not paying anything at all. Clearly, apart from Dunzo, no one else should have access to make such changes, and which is why setting access control matters. Why Setting the Right Access Control in a Smart Contract Matters Now, let’s deploy this smart contract and see how easy it is for an intruder to alter the values in the state variables. With both the setPrice and setPackageDelivered functions accessible, anyone can make changes that can result in financial loss for Dunzo delivery. If Dunzo had set the original price of the item at 5 ETH, the customer can now alter that value to 3 ETH. Of course, since the customer can also access the setPackageDelivered function and change the boolean value to false, he might receive another delivery of the same item. So, in either case, Dunzo stands to lose money, and might not even realize it until it is too late. For example, let’s say that the product being delivered is worth 5 ETH, as shown below: Upon deployment, no one but Dunzo should be able to modify contract terms. Still, when we switch addresses in Remix and reset the price to 3, this is possible due to no protection in place. We can even change the setPackageDelivered status from true to false. Even if the package has already been delivered to the customer. As you can tell, it’s important to distinguish between the owner and other users for said smart contract. So, let’s look at 4 fixes that will set the proper access control for this smart contract and ensure its safety as a result. 4 Ways to Setting Access Controls Safely for Your Smart Contract So, how can one differentiate between the owner and other users when writing smart contracts? Clearly, this is necessary due to the financial loss that can be incurred if ignored. Method #1: Use a require statement As you can see in the code below, a new state variable owner of address type has been added along with a ‘require’ statement. In the constructor, the owner state variable stores the address that is used when deploying the contract. As you know constructors in smart contracts are invoked only once, so the address cannot be changed. Now, when you invoke the setPrice function with a lesser value and a different address, the transaction is reverted to the initial state, much like the message below: As you can see, the reason provided by the contract involves the error message that we added to the ‘require’ statement. That said, no one but the owner of the contract — Dunzo, in this case — can change the sales terms. Method #2: Use a function modifier As simple as it is to add a ‘require’ statement with the right condition, a modifier is a block of code that you can use repeatedly. This certainly makes much more sense than writing numerous ‘require’ statements for such a check. In the code shared earlier, only the setPrice function is protected by a ‘require’ statement but nothing has been done for the setPackageDelivered function here. Using a modifier should do the trick, as shown in the Solidity code below: Now, if you try to access either the setPrice or setPackageDelivered functions, you’ll get the same error message obtained earlier, and as shown below: Method #3: Ownable This fix requires you to use the Ownable smart contract provided by OpenZeppelin. First, you have to import the Ownable smart contract and then add the onlyOwner modifier to the functions that you would like to protect. So, as shown below, both the setPrice and setPackageDelivered functions have the onlyOwner modifier below. Now, since the dunzoDelivery smart contract will use some of the functions in Ownable.sol, the “is Ownable” is added to the contract name too. That said, there are two other functions that one can access when using the Ownable smart contract: renounceOwnership and transferOwnership. While the account that was used to deploy the contract is usually considered to be the owner, this can be changed using these functions. Apart from seeing the owner’s address and the other agreement details, you can see the renounceOwnership and transferOwnership functions for use below: As expected, when you try to invoke the setPrice function using another address, the transaction will not go through. Instead, it reverts to the original state, with the reason provided below: Using the Ownable smart contract is good if you require only a single owner among other users. If you’re looking for more hierarchy, then the AccessControl.sol smart contract should be used. Method #4: AccessControl This last fix requires you to access the AccessControl smart contract by Open Zeppelin. For starters, you have to add the Access Control hyperlink to the import statement as well as add “is AccessControl” to the contract statement. As mentioned earlier, this smart contract allows you to add a number of roles within a hierarchy and which you must declare, as in the first two lines of the smart contract shown below: In this contract, there are two roles namely Dunzo and Customer. Now, when you deploy the contract, you will have to assign the aforementioned declared roles to said addresses, so as to determine who can access what. In addition, two ‘require’ statements have been added to the setPrice and setPackageDelivered functions. As you can tell, only the account that has been assigned the Dunzo role will be able to alter the sales terms mentioned during the time of deployment but not any other. That said, you can assign a number of roles with differing levels of access to functions of a smart contract using AccessControl.sol while Ownable.sol will not go into that much depth. A Note on Other Access Control Fixes for Ethereum Smart Contracts Now, these are not the only solutions available to implement access control in smart contracts. and have also been available in the past and which can associate roles with addresses too, much like AccessControl.sol and Ownable.sol. RBAC.sol WhitelistCrowdSale.sol So, if you want to gain a thorough understanding as to the evolution of implementing access control, it will be well worth your time, going over code in these two smart contracts too. Also published here. The lead image for this article was generated by via the prompt "access denied biometric scan". HackerNoon's AI Image Generator

How to Set Access Control for Smart Contracts

স্মার্ট চুক্তির জন্য কীভাবে অ্যাক্সেস নিয়ন্ত্রণ সেট করবেন

So richten Sie die Zugriffskontrolle für Smart Contracts ein

Cómo configurar el control de acceso para contratos inteligentes

Comment définir le contrôle d'accès pour les contrats intelligents

स्मार्ट कॉन्ट्रैक्ट्स के लिए एक्सेस कंट्रोल कैसे सेट करें

スマート コントラクトのアクセス制御を設定する方法

스마트 계약에 대한 액세스 제어를 설정하는 방법

Como definir o controle de acesso para contratos inteligentes

Akıllı Sözleşmeler için Erişim Kontrolü Nasıl Ayarlanır

Cách đặt quyền kiểm soát truy cập cho hợp đồng thông minh

如何设置智能合约的访问控制

Just like helps people share money between bank accounts seamlessly, the new Account Aggregator framework (AA) will help people share their financial data quickly and safely. UPI In the early days of UPI, people felt a lot of apprehensions. Is UPI safe? Is it okay to link my bank account to an app like ? How will the app be able to make transactions for me? How can I make sure that I don't make a false transaction? BHIM Since then, most of us have learned our way through UPI without falling prey to UPI scams and frauds. However, as AA apps start launching, we will have the same apprehensions about the AA ecosystem. Will the AA apps be safe? Will I be able to share my data securely? This post helps debunk myths around Account Aggregators so you can feel safe adopting these new apps. What is an Account Aggregator? Just like how a UPI app facilitates the transfer of funds from one bank to another, an Account Aggregator will be used to transfer financial data. Here's what you can do with an AA: Discover and link your accounts with . FIPs are the organisations that hold your financial data — e.g banks, insurance companies, mutual funds, pension funds, etc. Financial Information Providers (FIPs) Give consent to share your information with a . FIUs are organisations that consume financial data to provide consumer services — e.g. lending agencies, insurance companies, personal wealth management companies, etc. Financial Information User (FIUs) See the history of your data sharing and revoke any consent. Let's go through an example. Suppose you need to send your bank statements to your insurance agency. Today, you have to go to your bank branch or online banking, download your bank statements, attest them, and deliver or send them to the insurance agency. But with Account Aggregators, everything will happen digitally. An AA app is like Dunzo, picking up something for you and delivering it to another place. You will open your AA app and link your bank account. This lets the AA, aka your Dunzo driver, access and fetch your banking data. For a Dunzo delivery, since all orders and communication happen within the app, everyone involved knows that your Dunzo driver is legit. Similarly, since the entire data transaction happens with the AA protocol, the bank will know that your request is valid. Next, the AA app will fetch your account information from the bank in an encrypted form. Just like Dunzo can't look inside a package you send, an AA will be data-blind. It will know the details of your transaction, but it will not see any of the data that you are sending. Next, the AA will deliver your data to the insurance agency. There's no need to follow up or send clarifications about your data. The insurance agency will be given all the information it needs — the consent to access your bank statements, how long the consent lasts for, how many months of statements you've shared, whether all your statements are available, etc. Just like for a Dunzo delivery, once your bank statements have been delivered safely, both you and your bank will receive a notification. Learn more about Account Aggregators : how they work, how they were created, and how they'll transform the fintech industry. : If I link my financial account to an AA, it will store my information. Myth #1 Think of an Account Aggregator as a postman. Its job is to know that you have asked it to deliver a sealed envelope from Point A (an FIP) to Point B (an FIU). Just like how a postman can't see inside your envelope, an AA can't see the details of your request. AAs are data-blind. At most, they will know that you are sharing certain information provided by an FIP to an FIU. They may temporarily store your encrypted information in their servers. (This allows them to aggregate information from other FIPs and then send it all to FIUs, much like how a postman will collect various letters and then deliver all of them to you at once.) But they can't view that information. Linking your account to AA is quite safe. They only save information about the FIPs linked to your account to help you transact easier. They will never have access to your actual bank statements, loan bonds, tax returns, etc. : If I link too many of my financial accounts to an AA, they will be able to profile me. Myth #2 Again, this isn't possible. Account Aggregators are data-blind, and they can't read the data you share using it. Their job is to collect data requests for you and share information on your behalf, like a postman. No matter many accounts you link, an AA won't be able to see the details of your bank accounts or any other financial information, which would be necessary to profile you. : AA apps can share my data without my consent. Myth #3 Account Aggregation is based on a tightly thought-out consent architecture whose core tenant is data privacy and security. AAs will require explicit consent on every data transaction that you do. AAs won't be able to download data from an FIP without an explicit consent request being raised or approved by you. And, just like you have to approve a UPI transaction in the BHIM app, you will have to enter your secure AA PIN on an AA app to approve every transaction. Also, as per the (which lays out rules for AAs), Account Aggregators can't be in any other business. They can only be an AA. This prevents an AA from establishing a side business and selling your encrypted data. Reserve Bank of India's Master Directive Mock app image for setting a user PIN on an Account Aggregator. Created by Parallel . : If I give consent once, an AA will have access to my data forever. Myth #4 Every consent will only be valid for as long as you want. There are two types of consent: one-time consent and recurring consent. In a , your data will only be shared once. one-time consent In a , your consent will only be valid for a certain amount of time (which you set). recurring consent Let's say I want to share my bank statements. I can choose to share the last three months of bank statements once, or I can share bank statements with an FIU once a month for the next three months. When and how your data is shared totally depends on the request an FIU makes and whether you consent to that request. Moreover, you will always have access to your consent logs, and you can revoke consent anytime. For example, even if you had given consent for three months, you can revoke the consent after two months. After all, it's your data. : There will be far too many AA apps, just like all the UPI apps ( , , , etc.) Myth #5 BHIM Google Pay PhonePe Yes and no. Yes, there will be multiple Account Aggregator apps in the market. In fact, right now, there are already . seven registered AAs However, not everyone can become an AA. Companies must get a license from the RBI to operate as an AA, which requires a thorough . ( ) application process Check out the application form here. Currently, consent history can't be shared across AA apps. However, if better data portability and sharing was added as a feature, it would mean that people could use just one AA app of their choice and keep all financial records in one place. : Using AAs will lead to more data theft, since bank statements can be shared in a few clicks. Myth #6 Actually, your financial information is more at risk in the non-AA world. How many times have you printed your bank statements, got them attested, and then gave them to some other institution? There are more chances for misuse — a bank employee can take photos or make photocopies and sell your data, or you might forget your bank documents in a public place. With Account Aggregators, you are giving explicit digital consent for data sharing. It's like moving from paying with cash to a UPI app — no chance of someone stealing your cash anymore! Moreover, with AAs, you will always have a written record of who is using your financial data and for what purpose. Mock app image for giving consent to share financial data on an Account Aggregator. Created by Parallel . : With so much data sharing, I won't be able to track who is using my data and why. Myth #7 This is a half-myth. Yes, Account Aggregators will increase data sharing. That's the whole purpose — to help you share your confidential financial information securely and privately. However, AAs make it possible to track every data transaction that you make, which is difficult to track in the non-AA world. : Someone could see all my financial details if they open my AA app. Myth #8 Security is at the heart of the AA ecosystem, so AA apps won't expose all your financial details. When you link your financial accounts with AA, even the AA app can't see your entire account number. The data is shared with the AA app as masked information. Even if you leave your phone open by mistake, and someone looks at all your accounts, all they will be able to see the last four digits of your account number. And they won't be able to do anything with your accounts unless they enter your secure AA PIN. The TL;DR version To sum it up quickly... An Account Aggregator is like a postman. It can send your financial data, but it will never be able to read or store it. You'll have to consent to every data transaction, so AAs can't share data without your say-so. You'll always know when your financial data is shared and with whom, and you can revoke consent at any time. AAs make your financial data far safer, even though it's quick and easier for you to share your data. All financial data is encrypted and masked, and even people who have your phone can't view it. An AA app can't view or use your data to profile you, and they can never go into a side business with your data. The RBI is limiting AA apps, which will be able to share your consent history from app to app, so there's no need for more than one AA app on your phone. We hope this helps to debunk some of the common AA myths out there and gives you a glimpse into how the new Account Aggregator framework will work. This article was written by Richa Verma from Parallel and was previously published here .

Introduction to Account Aggregators: 8 Common Myths  Debunked

Account Aggregators (AAs) are about to take the fintech market by storm. Remember how UPI completely changed how people send money? No more awkwardly splitting restaurant bills, or digging for exact change to pay off that friend who's been bugging you for months. Account Aggregators (AAs) are about to do that for all other types of financial transactions in India! Imagine sending your bank statements in a single click, rather than having to print and deliver them. Or small businesses being able to share their financial data and get a loan in minutes, not months. Or sharing valid financial documents without having to sign them 100 times and show 10  forms of ID. Whether you're a fintech startup or a financial giant, it's imperative to learn the ins and outs of AAs before they transform the Indian fintech industry. Looking for a primer on what an Account Aggregator is? Or how your fintech company can use it? Or where it should fit into your fintech product? Or what DEPA, OCEN, and the rest of the AA jargon actually means? Keep reading for an explanation of all Account Aggregator concepts in layman's terms, a clear understanding of the changing fintech landscape, and how to prepare your product infrastructure and internal capabilities to adapt to AAs. What Exactly Is an Account Aggregator? Account Aggregators share data securely between FIPs and FIUs. : organisations that hold your financial data e.g banks, insurance companies, mutual funds, pension funds, etc. Financial Information Providers (FIPs) organisations that consume financial data to provide consumer services e.g. banks, lending agencies, insurance companies, personal wealth management companies, etc. Financial Information User (FIUs): An Account Aggregator will facilitate the process of consent, much like how a UPI app facilitates the transfer of funds from one bank account to another. AAs will be used to transfer financial data rather than funds. : Think of an AA as Dunzo for financial data. Example Let's work through an example: suppose you need to send your bank statements to your insurance agency. Today, you have to go to your bank branch or online banking, download your bank statements, attest them, and deliver or send them to the insurance agency. With Account Aggregators, everything will happen digitally. An AA app is like Dunzo, picking up something for you and delivering it to another place. You will open your AA app and link your bank account. This lets the AA, aka your Dunzo driver, access and fetch your banking data. For a Dunzo delivery, since all orders and communication happen within the app, everyone involved knows that your Dunzo driver is legit. Similarly, since the entire data transaction happens with the AA protocol, the bank will know that your request is valid. Next, the AA app will fetch your account information from the bank in an encrypted form. Just like Dunzo can't look inside a package you send, an AA will be data-blind. It will know the details of your transaction, but it will not see any of the data that you are sending. Next, the AA will deliver your data to the insurance agency. There's no need to follow up or send clarifications about your data. The insurance agency will be given all the information it needs — the consent to access your bank statements, how long the consent lasts for, how many months of statements you've shared, whether all your statements are available, etc. Just like for a Dunzo delivery, once your bank statements have been delivered safely, both you and your bank will receive a notification. And, much like Dunzo, the AA app will maintain logs of all your transactions, or in the AA world, all consents given and fulfilled. Mock user flow for giving consent to share data via an Account Aggregator. Created by Parallel . A Quick History of Account Aggregators 2016: RBI's Master Directive The Account Aggregator framework was introduced to make sharing financial data easier, quicker and more secure. This idea first showed up in government policy in 2016, when the Reserve Bank of India issued a to introduce how AAs would work and be regulated. Master Directive This framework was created by an inter-regulatory alphabet soup of : the Reserve Bank of India (RBI), Securities and Exchange Board of India (SEBI), Insurance Regulatory and Development Authority (IRDAI), and Pension Fund Regulatory and Development Authority (PFRDA) through the Financial Stability and Development Council (FSDC). government agencies AAs are the final layer (called the " ") of , a set of APIs to improve digital services for Indians. The other layers of India Stack include , , , and . Consent Layer India Stack Aadhar eKYC UPI DigiLocker eSign 2020: OCEN and DEPA In the last year, AAs have quickly moved from idea to an imminent reality, thanks to OCEN and DEPA. In July 2020, the was launched to help small businesses get credit easily. OCEN is a common language to help financial lenders and marketplaces communicate and create massive financial credit products. Open Credit Enablement Network ( OCEN ) "The Unified Payments Interface (UPI) was a protocol for payments, OCEN allows you to connect lenders to marketplaces." – Nandan Nilekani Account Aggregators are the biggest implementation of data sharing that can happen through OCEN. They will help lenders lend money to small businesses and less data-savvy people in minutes, not months. In addition, the will be launched in 2020. It has been by the NITI Aayog and is currently under discussion by both government and private stakeholders. Data Empowerment and Protection Architecture ( DEPA ) drafted DEPA is a framework that will help users safely, privately share their digital data. Its technology architecture allows for the creation of , data-blind tools that will safely share but never see, encrypted data. All data sharing requires a user's full consent and knowledge. In the financial sector, Consent Managers are called Account Aggregators. Consent Managers AA Apps Come to Your Phone (Soon) The fintech ecosystem is about to change dramatically as more and more fintech players join the Account Aggregator framework, DEPA and OCEN are implemented, and the RBI's Master Directive comes into effect. So far, have been approved by the RBI and are in the works. seven AA clients This is certainly just the start of how Account Aggregators will transform financial products. These new frameworks and policy measures will change the way the fintech industry functions in India, and potentially even set examples for the global fintech industry as well. Why are Account Aggregators important? "Today of Indian small businesses get credit from the banking system. Due to combination of factors like public sector banks having high NPAs, NBFCs having asset liability mismatch, the whole thing is gridlocked." – only 8% Nandan Nilekani Today, most small businesses find it difficult to get credit. They either lack the right data, or they don't have time to collect all the data they need. Using the AA architecture, banks and lending agencies will now be able to give credit based on verified data such as GST invoices, bank statements, securities information, etc. No need for repeated checks or physical examination of the data. Since the data will be coming in from the source itself, there are fewer chances of tampering with the data, and the access to this data will be quite easy. Easy sharing of verified data will truly digitise the lending process. It will make lending simpler for banks and businesses alike, and it will help banks give credit faster and more securely. Worried about whether Account Aggregators are safe? Check out our blog with 8 common myths about AAs. Mock user flow for linking your bank account on an Account Aggregator. Created by Parallel . How Will Account Aggregators Impact Fintechs? As Account Aggregators become more common, the fintech landscape will change drastically. Standardised Financial Information As the AA architecture is implemented, it will help standardise financial information across the fintech players. After all, data can't be easily shared if different organisations store it in different formats. Sharing and storing information in the same structure will make it quicker to transfer information and easier to create underwriting algorithms or other monitoring mechanisms. Best Practices for Security and Privacy The Account Aggregator framework has strong rules around data privacy, protection and access. Any organisation participating in AAs will have to follow these policies for trustworthy data sharing. As a result, AAs will encourage organisations to follow standardised security and privacy best practices e.g. data masking, encryption at source, and most importantly, explicit consent from a user before their data is shared or accessed. Safer Data Storage and Sharing Similarly, AA policies will mean no more ad-hoc ways of storing, maintaining and sharing data. Organisations will have to follow stricter security and privacy rules to share their data through AAs. And, most importantly, AAs will remove the risk of leaking financial information by printing it and sharing at branch offices or downloading it as PDFs and sharing it on emails, etc. Integrations for Current Financial Players As Account Aggregators becomes common, organisations will start to explore how to integrate the AA architecture and make their systems compliant. For example, if you're a lending institution, you could look at how to incorporate transaction-based lending, since AAs will make it easy to receive a user's transaction data. How might your underwriting algorithms change? How can you better monitor and track whether a lendee will pay you back? New Players in Fintech With AAs, it will be super quick to receive and verify someone's official bank statements. This makes it easier to onboard a new customer and verify their financial information without much time or resources. This will lead to the growth of new players in fintech sub-sectors e.g. lending, wealth management, personal finance management, (automated financial advice), and even accounting. robo-advisory What will Account Aggregators Mean for Fintech Customers? Higher Trust Implementing Account Aggregators will help fintech organisations build more trust with their customers. First, by making consent explicit, customers will be able to quickly check what information about them an organisation has or doesn't have. Second, given the security and privacy policies implemented because of AA, there will be a higher degree of trust when sharing financial information. Easier Data Collection AA will also reduce the pain of sharing information — finding financial information, collating it, logging it multiple times in various applications, and then sharing it securely with an FIU. For example, just for a loan, a customer has to separately collect six months of bank statements, income tax information, identification proofs, and more, all on paper. However, once AA is implemented, all information sharing can happen online, people won't need lots of identification proofs, and it won't take weeks to process an application because of the sheer amount of paperwork and verification. This means not just more privacy and security, but also better customer experience. Clear Data Consent and Logs The biggest upside with Account Aggregators is that consumers will always have the logs of all their data transactions. And, more importantly, they can revoke access whenever they want. It's not like handing over your documents to an organisation forever. Customers can revoke access and take their documents back without visiting a single office. Conclusion These are interesting times for the finance industry, not just in India, but across the world. Successful implementation of the DEPA framework and AA architecture will be instrumental for setting gold standards of protecting data, building consent into data transactions, and making data interoperability a reality. Next steps for Account Aggregator integrations 1. Consider how you can integrate AAs into your product or organisation: First, identify whether you will be aN FIP, an FIU, or both, and then prepare your organisation for the incoming change. Second, look up the (e.g. FinVu, OneMoney, etc.) and see how you can do a quick implementation or POC. upcoming AAs 2. Read the AA . API sandbox documentation ‍ 3. Check out the that AAs will be able to get from financial information providers (FIPs) and share with financial information users (FIUs). schema of information This article was written by Richa Verma and Christine Garcia from Parallel and was previously published here: https://www.parallelhq.com/blog/what-are-account-aggregators-fintech

Account Aggregators: What Are They and How Will They Change the Fintech Ecosystem?

Charles Darwin once said, "It is not the most intellectual of the species that survives; it is not the strongest that survives, but the species that survives is the one that can adapt and adjust best to the changing environment in which it finds itself. The global pandemic has caused upheaval across nations and pushed office workers indoors. Seen through a Darwinian lens, COVID-19 is forcing enterprises to adapt to new remote working conditions and technologies. It has forced three years of digital transformation to be fast-forwarded into a space of five months. Remote workers using unsecured devices and networks, and the digitization now integrated into our day-to-day work processes has made the business environment mobile and seamless. Data that is especially vulnerable to cyberattacks is now scattered everywhere. Dunzo’s data breach, which leaked phone numbers and email addresses of its users after the servers of a third-party were compromised, is a recent example. It allowed the attackers unauthorized access to the company database. Fortunately, no payment information, like credit or debit cards, etc. was leaked. It was, nevertheless, a significant breach of personal information. Sectors handling large amounts of data - like IT, banking, infrastructed and, recently, healthcare - are preferred targets for cybercriminals. India has seen a 37% rise in cyberattacks in the first quarter of 2020. According to the Kaspersky Security Network report, its products have detected and blocked around 52,820,874 local cyber threats between January to March 2020. Gone are the days when firewalls provided adequate data security against external threats. Data is no longer safe inside corporate strongholds and bad guys no longer need to penetrate fortified, high-tech perimeters to steal data. With weaker boundaries protecting corporate data, threats are just as likely to come from inside an organization's walls. According to , 95% of all corporate involve human error due to a lack of education and understanding of what makes good cyber hygiene. The heavy reliance on remote work has increased the risk of cyberattacks from phishing emails and malicious plugins. Amid the uncertainty about the lifespan of COVID-19, remote working will be a new normal necessitating a new security model. Enter: the zero trust security model. IBM research What is the zero trust security model? Zero trust security is a cybersecurity strategy based on the concept that an organization should not automatically trust anything, whether it be inside or outside its firewalls or browser isolation systems. In this particular security model, there is an assumption that the threat is already present inside the system and grants access only to authenticated users, devices, and applications to other users, devices, and applications within an organization. Technologies and strategies behind Zero Trust Security Here are the fundamental principles behind Zero Trust Security- 1. Microsegmentation- It involves breaking the networks into different security segments that are kept down to the individual level of workload to which the security controls about each workload are defined. 2. Multifactor Authentication – Users are granted network access on showing two or more forms of identification or credentials based on knowing passwords, fingerprint or retina scan, or having a token number. 3. Identity Access Management- Identity Access Management (IAM ) is a combination of features like Multifactor Authentication and Single Sign-On that enables users to securely authenticate with multiple applications and websites by logging only once. 4. Analytics- Implementing Zero trust security involves the analysis of data to detect abnormalities in user and device behaviour data that generate risk scores. The higher the risk scores means further verification is needed. 5. Orchestration- Orchestration aligns your processes for rapid response, replaces slow manual functions with automation, commands action to the needed enforcement points, and consolidates your entire security operation. 6. Encryption- This technique converts a sensitive data into a code that is difficult to decipher to prevent unauthorized use. Assuming that threats can be inside or outside, it will be prudent, to code the internal communications and passwords of a company in case there is a breach. Cybercriminals might attack the encrypted data through key access since efficient key management plays a pivotal role in the successful running of zero-trust security. 7. File system Permission – This technique controls the access of the user ability to view, navigate, change, or execute the contents of a protected file system. Functions may be made available or hidden on a user permission level. Businesses that want to survive and successfully operate in the decentralized working condition needs to think of its cybersecurity on a higher pedestal than what it has been in the past. Implementing a zero-trust security model can enable remote and agile ways of working, without the fear of stealing corporate data from both fronts internally and externally.

Zero-Trust Architecture Security: Decoded

I know I know what y’all gonna say. Give us a break you Bitcoin ignorant bald eagle; we’ve heard the same old story a thousand times. I heard it too. And, I ain’t talking about Bitcoin Pizza Day. The rumor has it that back in the early days, there were people who said no to Bitcoin at some presentation for FREE! Here’s the thing. It feels completely different to read about these stories and actually hear one first-hand in person. I can tell you; it can be very unpleasant and even painful. So, my story goes, this time of the year, I’m done with the summer carpet cleaning job, and looking for something else more warm and secure to get me through the winter. You know, carpet cleaning by day, stories writing by night. So, here I am, at my latest but not last job interview. How do you feel about a delivery job? What can you tell me about your driving skills? You know me; I’m honest to the bone. I dunno why my never-to-be-boss hadn’t shown me the door right away. Maybe he just wanted to kill some time before talking to the next candidate or maybe he found me interesting. Never mind. Moving on with my story. It turned out, he was quite a character. He used to live and work in Australia. You know how it goes. He married a lovely Oz lady and ran a successful business. Life was good for his family with two boys. He was also passionate about investing. The oil was his thing. Nowadays, $10K for investing may not seem as much, despite the economic sh*tstorm, but more than 10 years ago, it was a respectable amount of money. So, one fine day, he clearly remembers it was Friday, his broker had a brilliant idea to switch to corn for a change. He wanted to stick to oil, but this silver tongue broker convinced him somehow to forget about the liquid gold for a moment. Later that day, the news broke out that the price of corn would go down. But, it was too late. He had a miserable weekend. On Monday morning, he’d lost half of his money. What to do with the remaining $5K? The hero of this story had also a thing for tech and innovation. Because the Internet wasn’t still the thing back then as it is now, he was going through some magazines at home. He stumbled upon an article about this new thing called Bitcoin, which he found to be very interesting. He would never forget that the price was, at that time, ELEVEN CENTS. So, he was thinking out loud. What if I invest one thousand dollars in Bitcoin?! His wife heard him. After you just lost five thousand dollars, do you want to lose one thousand more? But, honey, it’s not that much, and look at the price, it’s only eleven cents. You’d think this was the moment where both of our stories end, but little did they know, it was only the beginning of a new chapter in their lives. Elysium “Med-Bay” Made in Serbia I told you already that this intriguing guy was a tech enthusiast. Fast forward a few years, he moved his entire family back to Belgrade, Serbia. This picture is worth a couple of thousand miles, give it or take. Here comes the tricky part, but I’m going to simplify it as much as possible. Have you seen the movie Elysium? Definitely not the finest product by director Neill Blomkamp compared to “District 9” and “Chappie,” but worth watching. The most interesting Elysium’s “gadget” is the so-called Med-Bay. It looks like a solarium (tanning) bed, but it’s not supposed to improve your looks. It’s all about improving your overall health. https://www.youtube.com/watch?v=XwaljThPDh0&embedable=true Mr. M was a tech startup entrepreneur before this term was even invented. As you can imagine, it’s been a road full of thorns and spikes. Production. Certification. Bureaucracy. COVID. You name it, he got hit by it. https://www.youtube.com/watch?v=xcYmlfhShIg&embedable=true He wasn’t giving up but at a high cost. They had to sell their house in Australia. He got stuck after spending more than $2.5M! The vans used for making his dreams come true are now a part of delivery business plan B. Yeah, it’s heartbreaking. I showed him my latest Bitcoin story. He liked it. Then, we had that “what if” moment. How many Bitcoins he could’ve bought for a thousand dollars when the price was eleven cents? How much money he would’ve had today at his disposal? Well, does it hurt like hell? Stupid people ask stupid questions. I was like, what about China? Is there a way out? China 2.0 - Tech on Steroids Things have changed, my friend. By the time I somehow managed to catch my breath after I don’t know how many years, guess what? I was all ears. The Chinese flooded the market with a similar product for half the price. Then, he shared a story about a fellow entrepreneur who was more lucky. Here comes the part of my story you have to read without prejudice because it backfires big time. His friend took a prototype and made a few calls to China. He didn’t go there, they came to him. There was an entire team of dozens of engineers. All of them with the tech degrees from the USA, Australia, or Canada. His friend was straightforward. So, what do you think? How long it’s going to take for you to make me more of this? The Chinese team just took a prototype and left, without a word. A few days later, they came back with two devices. His friend couldn’t tell the difference between his original prototype and the “made in China” model. Various articles in the popular media, speeches by policy-makers, and reports to Congress have stated that the United States graduates roughly 70,000 undergraduate engineers annually, whereas China graduates 600,000 and India 350,000. I also know this is old news, but it’s worth repeating and remembering: I wrote a story about Elon Musk’s plans for the city, Terminus, on Mars. I have no problems with Chinese being either the dominant or official language on Mars. Y’all heard that story about the Chinese version of TikTok with completely different content preferences and algorithms for kids compared to the rest of the world, haven’t you? It was about time for me to hit the road. On my way out, I had to ask. Have you ever said something to your wife about you know what? No. Never. He laughed. She turned out to be my only successful investing choice. We have two beautiful and smart sons. She stood by my side and supported me all these years. Can I write a story about you? Are you OK with it? Sure. Why not. Send me a link once it’s published, I want to read it with my wife. We both laughed. On the way to my mortgageous and sadly empty home, I couldn’t stop thinking how little it takes to be at the right time, at the right place, and do the best thing possible. You know how they say, after a battle, everyone is a general. Well, in 2024, everybody is a Bitcoin wise guy. It’s easy to judge, but it’s hard to predict. I mean, I don’t even know what’s going to happen with our world tomorrow. But still, eleven cents. Oh boy, the crazy unforgettable moments. This one goes to all bald eagles out there regardless of how much they know about Bitcoin: Disturbed - The Sound Of Silence. I know I know what y’all gonna say. Give us a break you Bitcoin ignorant bald eagle; we’ve heard the same old story a thousand times. I heard it too. And, I ain’t talking about Bitcoin Pizza Day . The rumor has it that back in the early days, there were people who said no to Bitcoin at some presentation for FREE! Bitcoin Pizza Day Here’s the thing. It feels completely different to read about these stories and actually hear one first-hand in person. I can tell you; it can be very unpleasant and even painful. So, my story goes, this time of the year, I’m done with the summer carpet cleaning job, and looking for something else more warm and secure to get me through the winter. You know, carpet cleaning by day, stories writing by night. So, here I am, at my latest but not last job interview. How do you feel about a delivery job? What can you tell me about your driving skills? You know me; I’m honest to the bone. I dunno why my never-to-be-boss hadn’t shown me the door right away. Maybe he just wanted to kill some time before talking to the next candidate or maybe he found me interesting. Never mind. Moving on with my story. It turned out, he was quite a character. He used to live and work in Australia. You know how it goes. He married a lovely Oz lady and ran a successful business. Life was good for his family with two boys. He was also passionate about investing. The oil was his thing. Nowadays, $10K for investing may not seem as much, despite the economic sh*tstorm, but more than 10 years ago, it was a respectable amount of money. So, one fine day, he clearly remembers it was Friday, his broker had a brilliant idea to switch to corn for a change. He wanted to stick to oil, but this silver tongue broker convinced him somehow to forget about the liquid gold for a moment. Later that day, the news broke out that the price of corn would go down. But, it was too late. He had a miserable weekend. On Monday morning, he’d lost half of his money. What to do with the remaining $5K? The hero of this story had also a thing for tech and innovation. Because the Internet wasn’t still the thing back then as it is now, he was going through some magazines at home. He stumbled upon an article about this new thing called Bitcoin, which he found to be very interesting. He would never forget that the price was, at that time, ELEVEN CENTS. So, he was thinking out loud. What if I invest one thousand dollars in Bitcoin?! What if I invest one thousand dollars in Bitcoin?! His wife heard him. After you just lost five thousand dollars, do you want to lose one thousand more? But, honey, it’s not that much, and look at the price, it’s only eleven cents. You’d think this was the moment where both of our stories end, but little did they know, it was only the beginning of a new chapter in their lives. Elysium “Med-Bay” Made in Serbia I told you already that this intriguing guy was a tech enthusiast. Fast forward a few years, he moved his entire family back to Belgrade, Serbia. This picture is worth a couple of thousand miles, give it or take. Here comes the tricky part, but I’m going to simplify it as much as possible. Have you seen the movie Elysium ? Definitely not the finest product by director Neill Blomkamp compared to “District 9” and “Chappie,” but worth watching. The most interesting Elysium’s “gadget” is the so-called Med-Bay . It looks like a solarium (tanning) bed, but it’s not supposed to improve your looks. It’s all about improving your overall health. Elysium Med-Bay https://www.youtube.com/watch?v=XwaljThPDh0&embedable=true https://www.youtube.com/watch?v=XwaljThPDh0&embedable=true Mr. M was a tech startup entrepreneur before this term was even invented. As you can imagine, it’s been a road full of thorns and spikes. Production. Certification. Bureaucracy. COVID. You name it, he got hit by it. https://www.youtube.com/watch?v=xcYmlfhShIg&embedable=true https://www.youtube.com/watch?v=xcYmlfhShIg&embedable=true He wasn’t giving up but at a high cost. They had to sell their house in Australia. He got stuck after spending more than $2.5M! The vans used for making his dreams come true are now a part of delivery business plan B. Yeah, it’s heartbreaking. I showed him my latest Bitcoin story . He liked it. Then, we had that “what if” moment. my latest Bitcoin story How many Bitcoins he could’ve bought for a thousand dollars when the price was eleven cents? How much money he would’ve had today at his disposal? How many Bitcoins he could’ve bought for a thousand dollars when the price was eleven cents? How much money he would’ve had today at his disposal? Well, does it hurt like hell? Stupid people ask stupid questions. I was like, what about China? Is there a way out? China 2.0 - Tech on Steroids Things have changed, my friend. By the time I somehow managed to catch my breath after I don’t know how many years, guess what? I was all ears. The Chinese flooded the market with a similar product for half the price. Then, he shared a story about a fellow entrepreneur who was more lucky. Here comes the part of my story you have to read without prejudice because it backfires big time. His friend took a prototype and made a few calls to China. He didn’t go there, they came to him. There was an entire team of dozens of engineers. All of them with the tech degrees from the USA, Australia, or Canada. His friend was straightforward. So, what do you think? How long it’s going to take for you to make me more of this? The Chinese team just took a prototype and left, without a word. A few days later, they came back with two devices. His friend couldn’t tell the difference between his original prototype and the “made in China” model. Various articles in the popular media, speeches by policy-makers, and reports to Congress have stated that the United States graduates roughly 70,000 undergraduate engineers annually, whereas China graduates 600,000 and India 350,000. Various articles in the popular media, speeches by policy-makers, and reports to Congress have stated that the United States graduates roughly 70,000 undergraduate engineers annually, whereas China graduates 600,000 and India 350,000. China graduates 600,000 China graduates 600,000 I also know this is old news, but it’s worth repeating and remembering: I wrote a story about Elon Musk’s plans for the city, Terminus, on Mars . I have no problems with Chinese being either the dominant or official language on Mars. Y’all heard that story about the Chinese version of TikTok with completely different content preferences and algorithms for kids compared to the rest of the world, haven’t you? Elon Musk’s plans for the city, Terminus, on Mars story about the Chinese version of TikTok It was about time for me to hit the road. On my way out, I had to ask. Have you ever said something to your wife about you know what? No. Never. He laughed. She turned out to be my only successful investing choice. We have two beautiful and smart sons. She stood by my side and supported me all these years. Can I write a story about you? Are you OK with it? Sure. Why not. Send me a link once it’s published, I want to read it with my wife. We both laughed. On the way to my mortgageous and sadly empty home, I couldn’t stop thinking how little it takes to be at the right time, at the right place, and do the best thing possible. You know how they say, after a battle, everyone is a general. Well, in 2024, everybody is a Bitcoin wise guy. It’s easy to judge, but it’s hard to predict. I mean, I don’t even know what’s going to happen with our world tomorrow. But still, eleven cents. Oh boy, the crazy unforgettable moments. This one goes to all bald eagles out there regardless of how much they know about Bitcoin: Disturbed - The Sound Of Silence . Disturbed - The Sound Of Silence

When Bitcoin Was Eleven Cents

Toe Bitcoin elf sent was

Bitcoin on bir sent olanda

Da Bitcoin var elleve cent

Cuando Bitcoin valía once centavos

Cando Bitcoin tiña once centavos

Երբ Bitcoin-ը տասնմեկ ցենտ էր

ビットコインが11セントだった頃

Биткоин он бір цент болған кезде

Кога Биткоинот беше единаесет центи

Вақте ки Bitcoin ёздаҳ сент буд

Bitcoin o'n bir sent bo'lganida

当比特币价格为 11 美分时

15+ Awesome Stuff to Build and Sell With JavaScript (and Friends)

I Fine-Tuned an LLM With My Telegram Chat History. Here’s What I Learned

আমি আমার টেলিগ্রাম চ্যাট ইতিহাসের সাথে একটি এলএলএম ফাইন-টিউন করেছি। আমি যা শিখেছি তা এখানে

Ich habe einen LLm mit meinem Telegram-Chatverlauf optimiert. Das habe ich gelernt

Ajusté una película con mi historial de chat de Telegram. Esto es lo que aprendí

J'ai affiné un Llm avec mon historique de discussion Telegram. Voici ce que j'ai appris

मैंने अपने टेलीग्राम चैट इतिहास के साथ एक एलएलएम को बेहतर बनाया। यहाँ मैंने क्या सीखा है

私はテレグラムのチャット履歴を使ってLLMを微調整しました。そこで学んだことは次のとおりです

내 텔레그램 채팅 기록으로 Llm을 미세 조정했습니다. 내가 배운 내용은 다음과 같습니다.

Ajustei um Llm com meu histórico de bate-papo do Telegram. Aqui está o que aprendi

Я настроил Llm с помощью истории чата в Telegram. Вот что я узнал

Telegram Sohbet Geçmişimle Bir Llm'ye İnce Ayar Yaptım. İşte Öğrendiklerim

Tôi đã tinh chỉnh Llm bằng lịch sử trò chuyện trên Telegram của mình. Đây là những gì tôi đã học được

我利用 Telegram 聊天记录对 LLM 进行了微调。以下是我学到的东西

How to Start a Career in Product Marketing: 10 Expert Tips (2024)

Authors: (1) Renato P. dos Santos, CIAGE – Centre for Generative Artificial Intelligence in Cognition and Education. Table of Links Abstract and Introduction Materials And Methods Results and Analyses Prompts and generated texts Conceptualizing chemical reactions Deepening on understanding of chemical reactions Question about combustion Question about a graph of gases turning into water over time Question about the difference between atoms, molecules, and moles Deepening on the concept of mole Question about changing of state Question about an animated representation of water molecules undergoing phase changes Question about plasma, a state of matter Question about chemical bondings Question about illustration of chemical bonds Question about the essence of the type of chemical bonding Further analysis Conclusions Limitations of the study and possible future studies Author Contributions, Conflicts of interest, Acknowledgements, and References Limitations of the study and possible future studies Despite its inherent limitations, including its single case design and the potential for bias, the exploratory depth of the study uncovered hidden potential within these systems, even amid serious concerns about generalizability. Future research could include: • Refining the crafting of prompts. • Exploring new features of these and other GenAIbots being introduced with increasing frequency. • Establishing standardised benchmarks to evaluate and compare chatbots and AI systems' performance, accuracy, and reliability. • Conducting long-term studies to observe the evolution of chatbots' capabilities and their impact on user interactions over time. • Conducting research with real students in classroom settings and beyond to assess these AI systems' practical educational applications and challenges. • Investigating chatbots' learning and adaptation capabilities to individual user needs and preferences over time. • Researching the integration of multimodal inputs (e.g., text, voice, image) to enhance chatbot capabilities and user interaction experiences. When implementing GenAIbots in Chemistry learning, it is crucial to evaluate benefits and drawbacks judiciously, ensuring accurate information delivery and considering the implications of reduced human interaction. These concerns may be alleviated by integrating GenAIbots with other educational tools or activities promoting collaborative dialogue among learners. Author Contributions The author confirms sole responsibility for the following: study conception and design, data collection, analysis and interpretation of results, and manuscript preparation. Conflicts of interest There are no conflicts to declare. Acknowledgements The authors warmly acknowledge Melanie Swan for her invaluable suggestion, which led to the transition from using the term 'objects-to-think-with' to 'agents-to-thinkwith'. References Adiguzel, T., Kaya, M. H., & Cansu, F. K. (2023). Revolutionising education with AI: Exploring the transformative potential of ChatGPT. Contemporary Educational Technology, 15(3), ep429. https://doi.org/10.30935/cedtech/13152 Baidoo-Anu, D., & Owusu Ansah, L. (2023). Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4337484 Bardin, L. (1977). L’analyse de contenu. PUF - Presses Universitaires de France. Bitzenbauer, P. (2023). ChatGPT in physics education: A pilot study on easy-to-implement activities. Contemporary Educational Technology, 15(3), ep430. https://doi.org/10.30935/cedtech/13176 Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., Agarwal, S., Herbert-Voss, A., Krueger, G., Henighan, T., Child, R., Ramesh, A., Ziegler, D. M., Wu, J., Winter, C., … Amodei, D. (2020). Language Models are Few-Shot Learners. ArXiv, 2005.14165. http://arxiv.org/abs/2005.14165 Castro Nascimento, C. M., & Pimentel, A. S. (2023). Do Large Language Models Understand Chemistry? A Conversation with ChatGPT. Journal of Chemical Information and Modeling, 63(6), 1649–1655. https://doi.org/10.1021/acs.jcim.3c00285 David, E. (2023, September 21). Microsoft to add DALL-E 3 to Bing Chat. The Verge; Vox Media. https://www.theverge.com/2023/9/21/23873690/microsoft-new-ai-features-bingsearch-shopping-dall-e-3 Dewi, C. A., Pahriah, P., & Purmadi, A. (2021). The Urgency of Digital Literacy for Generation Z Students in Chemistry Learning. International Journal of Emerging Technologies in Learning, 16(11), 88–103. https://doi.org/10.3991/ijet.v16i11.19871 Dunlop, L., Hodgson, A., & Stubbs, J. E. (2020). Building capabilities in chemistry education: happiness and discomfort through philosophical dialogue in chemistry. Chemistry Education Research and Practice, 21(1), 438–451. https://doi.org/10.1039/C9RP00141G Flavell, J. H. (1976). Metacognitive aspects of problem-solving. In L. B. Resnick (Ed.), The Nature of Intelligence (pp. 231–236). Lawrence Erlbaum. Floridi, L., & Chiriatti, M. (2020). GPT-3: Its Nature, Scope, Limits, and Consequences. Minds and Machines, 30(4), 681–694. https://doi.org/10.1007/s11023-020-09548-1 Franciscu, S. (2023). ChatGPT: A Natural Language Generation Model for Chatbots. https://doi.org/10.13140/RG.2.2.24777.83044 Gregorcic, B., & Pendrill, A.-M. (2023). ChatGPT and the frustrated Socrates. Physics Education, 58(3), 035021. https://doi.org/10.1088/1361-6552/acc299 Latour, B. (1991). Nous n’avons jamais été modernes: Essai d’anthropologie symétrique. La Découverte. Leon, A. J., & Vidhani, D. (2023). ChatGPT Needs a Chemistry Tutor, Too. Journal of Chemical Education. https://doi.org/10.1021/acs.jchemed.3c00288 Liu, F., Budiu, R., Zhang, A., & Cionca, E. (2023, October 1). ChatGPT, Bard, or Bing Chat? Differences Among 3 Generative-AI Bots. https://www.nngroup.com/articles/ai-botcomparison/ Metz, C., & Hsu, T. (2023, September 20). ChatGPT Can Now Generate Images, Too. New York Times - Technology. https://www.nytimes.com/2023/09/20/technology/chatgpt-dalle3-images-openai.html Mishra, A., Soni, U., Arunkumar, A., Huang, J., Kwon, B. C., & Bryan, C. (2023). PromptAid: Prompt Exploration, Perturbation, Testing and Iteration using Visual Analytics for Large Language Models. ArXiv, 2304.01964. https://doi.org/10.48550/arXiv.2304.01964 Mollick, E. R. (2023, April 26). A guide to prompting AI (for what it is worth): A little bit of magic,but mostly just practice. One Useful Thing Blog. https://www.oneusefulthing.org/p/a-guidetoprompting-ai-for-what Okonkwo, C. W., & Ade-Ibijola, A. (2021). Chatbots applications in education: A systematic review. Computers and Education: Artificial Intelligence, 2, 100033. https://doi.org/10.1016/j.caeai.2021.100033 OpenAI. (2023). GPT-4 Technical Report. OpenAI. https://cdn.openai.com/papers/gpt-4.pdf Papert, S. A. (1980). Mindstorms - Children, Computers and Powerful Ideas. Basic Books. http://www.arvindguptatoys.com/arvindgupta/mindstorms.pdf Permatasari, M. B., Rahayu, S., & Dasna, W. (2022). Chemistry Learning Using Multiple Representations: A Systematic Literature Review. J.Sci.Learn.2022, 5(2), 334–341. https://doi.org/10.17509/jsl.v5i2.42656 Pimentel, A., Wagener, A., Silveira, E. F. da, Picciani, P., Salles, B., Follmer, C., & Jr., O. N. O. (2023). Challenging ChatGPT with Chemistry-Related Subjects. ChemRxiv. https://doi.org/10.26434/chemrxiv-2023-xl6w3 Schlosser, M. (2019). Agency. In The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/win2019/entries/agency/ Swan, M. (2015, January 21). We Should Consider the Future World as one of Multi-Species Intelligence. Response to The Edge Question 2015: What Do You Think About Machines That Think?; Edge.org. http://edge.org/response-detail/26070 Taylor, C. A., Hogarth, H., Hacking, E. B., & Bastos, E. (2022). Posthuman Object Pedagogies: Thinking with Things to Think with Theory for Innovative Educational Research. Cultural and Pedagogical Inquiry, 14(1), 206–221. https://doi.org/10.18733/cpi29662 Timilsena, N. P., Maharjan, K. B., & Devkota, K. M. (2022). Teachers’ And Students’ Experiences In Chemistry Learning Difficulties. Journal of Positive School Psychology, 6(10), 2856–2867. https://www.journalppw.com/index.php/jpsp/article/view/13764 Tümay, H. (2016). Reconsidering learning difficulties and misconceptions in chemistry: emergence in chemistry and its implications for chemical education. Chemistry Education Research and Practice, 17(2), 229–245. https://doi.org/10.1039/C6RP00008H Turkle, S. (1984). The Second Self: Computers and the Human Spirit. Simon and Schuster. Yin, R. K. (2011). Applications of Case Study Research. SAGE Publications, Inc. This paper is under CC BY-SA 4.0 DEED license. available on arxiv

Overcoming Limitations in AI Chatbot Research for Future Educational Impact

ভবিষ্যতের শিক্ষাগত প্রভাবের জন্য এআই চ্যাটবট গবেষণায় সীমাবদ্ধতা অতিক্রম করা

Überwindung von Einschränkungen in der KI-Chatbot-Forschung für zukünftige Bildungseffekte

Superar las limitaciones en la investigación de chatbots de IA para lograr un impacto educativo futuro

Surmonter les limites de la recherche sur les chatbots IA pour un impact éducatif futur

भविष्य के शैक्षिक प्रभाव के लिए AI चैटबॉट अनुसंधान में सीमाओं पर काबू पाना

将来の教育への影響のために AI チャットボット研究の限界を克服する

미래 교육적 영향을 위한 AI 챗봇 연구의 한계 극복

Superando as limitações na pesquisa do AI Chatbot para um impacto educacional futuro

Преодоление ограничений в исследованиях чат-ботов с использованием искусственного интеллекта для будущего воздействия на образование

Gelecekteki Eğitim Etkisi için Yapay Zeka Chatbot Araştırmalarında Sınırlamaların Aşılması

Khắc phục những hạn chế trong nghiên cứu Chatbot AI để tạo ra tác động giáo dục trong tương lai

克服人工智能聊天机器人研究的局限性，为未来教育带来影响

Stories, News & Mentions about DUNZO

Dunzo’s Stories on HackerNoon

Dunzo's News & Mentions