Aecom (ACM) receives a strong valuation score of 78 from InvestorsObserver

Aecom (ACM) Stock: How Does it Score on Valuation Metrics?

Infrastructure Ontario (IO) and transit agency Metrolinx have selected the ‘Pape North Connect’ team to develop the Ontario Line subway project’s Pape

Arcadis and AECOM to design Ontario Line tunnel and underground stations - Canadian Consulting Engineer

TORONTO — Aecom has announced that Aleksandar Zivkovic has been named mining market sector lead for its Canadian water business. Zivkovic’s career includes design and construction projects where he first specialized in site selection, engineering geology and rock engineering for hydropower ...

Aecom appoints Zivkovic mining lead for Canadian water business - constructconnect.com

The 30 percent design, created by AECOM, is expected to be finished in the early spring

City, AECOM to host meetings with property owners along Bus. 51 - Point/Plover Metro Wire

The departures add six venue designers since last May to the existing Gensler practice.

Sports Architects Jump From AECOM to Gensler | Engineering News-Record

A Colorado federal court will allow a contractor to prove up more than $250 million in damages using the modified total cost method (see AECOM Technical Services v. FlatIron AECOM, LLC, Case No. 19-CV-2811, 2024 WL 22640 (D. Co. Jan. 2, 2024)). The AECOM v.

AECOM Technical Services v FlatIron AECOM LLC Decision

AECOM , the world's trusted infrastructure consulting firm, today announced that it has been selected by Metro Vancouver to provide engineering services for the design and construction of the Stoney Creek Trunk Sewer Upgrade project.

AECOM selected to provide engineering services for Metro Vancouver's Stoney Creek Trunk Sewer Upgrade

CORE Construction, Gilbane, Balfour Beatty, Skanska USA, and Adolfson & Peterson top BD+C's ranking of the nation's largest K-12 school building contractors and construction management (CM) firms for 2023,

Top 100 K-12 School Construction Firms for 2023

AECOM, CMTA, Jacobs, WSP, and IMEG head BD+C's ranking of the nation's largest K-12 school building engineering and engineering/architecture (EA) firms for 2023, as reported in Building Design+Construction's 2023 Giants 400 Report.

Top 80 K-12 School Engineering Firms for 2023

The Dallas-based firm was accused of securing more than $100 million in overpayments for Xavier University, Dillard University, The Archdiocese of New Orleans and several local schools.

Accused of cheating FEMA after Hurricane Katrina, Texas firm agrees to settle fraud case

AECOM

aecom

Reality Is Flexible: You Just Have to Believe

La realitat és flexible: només has de creure

La realidad es flexible: sólo tienes que creer

वास्तविकता लचीला है: आपको बस विश्वास करना है

Realnost je fleksibilna: samo trebate vjerovati

Realitas adalah fleksibel: Anda hanya perlu percaya

現実は柔軟である:あなたは信じなければならない

ការពិតគឺមានភាពងាយស្រួល: អ្នកត្រូវតែគិតថា

현실은 유연합니다 : 당신은 단지 믿어야합니다.

Verkligheten är flexibel: Du måste bara tro

Reality is Flexible: நீங்கள் மட்டுமே நம்ப வேண்டும்

حقیقت انعطاف پذیر ہے: آپ کو صرف یقین کرنے کی ضرورت ہے

Reality flexible: siz o‘z o‘z o‘z o‘z o‘z

Tech Stack Decisions are Make-or-break for Startups

Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 2 Related Work 2 Related Work 3 Method and 3.1 Phase 1: Taxonomy Generation 3 Method and 3.1 Phase 1: Taxonomy Generation 3.2 Phase 2: LLM-Augmented Text Classification 3.2 Phase 2: LLM-Augmented Text Classification 4 Evaluation Suite and 4.1 Phase 1 Evaluation Strategies 4 Evaluation Suite and 4.1 Phase 1 Evaluation Strategies 4.2 Phase 2 Evaluation Strategies 4.2 Phase 2 Evaluation Strategies 5 Experiments and 5.1 Data 5 Experiments and 5.1 Data 5.2 Taxonomy Generation 5.2 Taxonomy Generation 5.3 LLM-Augmented Text Classification 5.3 LLM-Augmented Text Classification 5.4 Summary of Findings and Suggestions 5.4 Summary of Findings and Suggestions 6 Discussion and Future Work, and References 6 Discussion and Future Work, and References A. Taxonomies A. Taxonomies B. Additional Results B. Additional Results C. Implementation Details C. Implementation Details D. Prompt Templates D. Prompt Templates 2 RELATED WORK Taxonomy Generation. Prior work in taxonomy generation falls into manual and automatic approaches. Handcrafted taxonomies, beyond being expensive to construct, tend to be developed for specific downstream tasks (e.g., web search intent analysis [4, 20], chatbot intent detection [31]), or tied to the development of specific datasets [22, 25]. On the other hand, automated approaches scale better but either rely on term extraction from corpora to obtain labels, which may hinder interpretability and/or coverage [24, 33], or else require a set of seeds for the taxonomy in order to generate new labels [32]. TnT-LLM, in contrast, is automatic, abstractive (i.e., labels describe the corpus but need not be directly extracted from it), and does not require any seed labels. Moreover, TnT-LLM treats taxonomy generation and text classification as interrelated problems in an end-to-end pipeline, whereas prior work has tended to focus mainly on the quality of the taxonomy produced, without considering its downstream utility for classification. Taxonomy Generation TnT-LLM TnT-LLM Text Clustering and Topic Modeling. Text clustering and topic modeling “invert” the traditional approach of defining a label set, then applying the labels on the corpus. Given a set of documents, such approaches first group the documents into topical clusters using various definitions of textual similarity, then post-hoc label or summarize the clusters [1, 29]. While traditional approaches in theory accomplish the same goals as TnT-LLM, they suffer due to a lack of interpretability [5], as they typically do not assign intelligible labels to clusters. More recently, attempts have been made to overcome these problems by using LLMs for topic modeling [18, 30], though these approaches still require supervision through either a predefined taxonomy [30] or a seed set of topics [18]. Text Clustering and Topic Modeling. TnT-LLM LLMs as Annotators. Recent work has explored using LLMs to replace human annotation for labor-intensive tasks such as search relevance quality labeling [28], topic and stance detection [8], and various computational social science labeling tasks [34]. These studies have found that, in general, LLMs perform on par or even better than crowd-workers [15], often at a fraction of the cost. In the same vein, we explore using LLMs as annotators for text classification, although our main goal is to scale the process by distilling LLMs’ label-specific capabilities into more efficient, lightweight classifiers. LLMs as Annotators. This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. This paper is available on arxiv under CC BY-NC-ND 4.0 DEED license. available on arxiv Authors:
(1) Mengting Wan, Microsoft Corporation and Microsoft Corporation;
(2) Tara Safavi (Corresponding authors), Microsoft Corporation;
(3) Sujay Kumar Jauhar, Microsoft Corporation;
(4) Yujin Kim, Microsoft Corporation;
(5) Scott Counts, Microsoft Corporation;
(6) Jennifer Neville, Microsoft Corporation;
(7) Siddharth Suri, Microsoft Corporation;
(8) Chirag Shah, University of Washington and Work done while working at Microsoft;
(9) Ryen W. White, Microsoft Corporation;
(10) Longqi Yang, Microsoft Corporation;
(11) Reid Andersen, Microsoft Corporation;
(12) Georg Buscher, Microsoft Corporation;
(13) Dhruv Joshi, Microsoft Corporation;
(14) Nagu Rangan, Microsoft Corporation. Authors: Authors: (1) Mengting Wan, Microsoft Corporation and Microsoft Corporation; (2) Tara Safavi (Corresponding authors), Microsoft Corporation; (3) Sujay Kumar Jauhar, Microsoft Corporation; (4) Yujin Kim, Microsoft Corporation; (5) Scott Counts, Microsoft Corporation; (6) Jennifer Neville, Microsoft Corporation; (7) Siddharth Suri, Microsoft Corporation; (8) Chirag Shah, University of Washington and Work done while working at Microsoft; (9) Ryen W. White, Microsoft Corporation; (10) Longqi Yang, Microsoft Corporation; (11) Reid Andersen, Microsoft Corporation; (12) Georg Buscher, Microsoft Corporation; (13) Dhruv Joshi, Microsoft Corporation; (14) Nagu Rangan, Microsoft Corporation.

Prior Approaches to Text Mining: Taxonomy, Clustering, and LLM Annotation

An AI Agent is the combination of a Large Language Model and a set of tools that the model uses for executing queries or accomplishing tasks in response to requests coming from Users or other Agents. AI agents have the potential to scale, automate, and improve business processes across various workplace functions and to significantly boost personal productivity. AI Agent Large Language Model There is a broad consensus that the “one agent fits all” approach is not suitable for the complexity of the tasks that Agents are expected to accomplish. The solution to this problem lies in Agentic Workflows, made by networks of autonomous AI Agents that make decisions, take actions and coordinate tasks with minimal human intervention. (IBM). Agentic Workflows, IBM IBM Google’s Proposal for Agent Interoperability: Agent2Agent Protocol (A2A) On April 9, 2025, Google announced the launch of the Agent2Agent (A2A)protocol, designed to enable AI Agents to communicate with one another, securely exchanging information and automating complex business workflows through interaction with enterprise and third-party platforms and applications. Agent2Agent (A2A)protocol The A2A protocol has been developed by Google in collaboration with more than 50 industry partners, who share a common vision of the future of AI Agent collaboration. This collaboration is independent from the underlying technologies and based on open and secure standards. A2A Design Principles As stated in the announcement, during the design of the A2A protocol, Google and its partners adhered to a few key principles: announcement announcement A2A must be an open, vendor-agnostic protocol, whose specifications are made publicly available;
A2A must allow agents to collaborate in their natural, unstructured modalities, hence drawing a clear distinction between Agents and Tools and distinguishing itself from the Model Context Protocol (see below);
The protocol must be built on existing standards, such as HTTP, SSE, and JSON-RPC, in order to be easy to integrate with existing IT stacks;
A2A must be secure by default, supporting enterprise-grade authentication and authorization;
A2A must be agnostic to the data modality, being able to handle text as well as images, audio and video streaming. A2A must be an open, vendor-agnostic protocol, whose specifications are made publicly available; open, vendor-agnostic protocol A2A must allow agents to collaborate in their natural, unstructured modalities, hence drawing a clear distinction between Agents and Tools and distinguishing itself from the Model Context Protocol (see below); collaborate in their natural, unstructured modalities Model Context Protocol Model Context Protocol The protocol must be built on existing standards, such as HTTP, SSE, and JSON-RPC, in order to be easy to integrate with existing IT stacks; built on existing standards A2A must be secure by default, supporting enterprise-grade authentication and authorization; secure by default A2A must be agnostic to the data modality, being able to handle text as well as images, audio and video streaming. agnostic to the data modality, What A2A Provides A2A provides the following functionalities out of the box: Capability discovery: The protocol allows the Agents to advertise their capabilities, helping the clients identify which agent is best for performing a task;
Task and state management: The interaction between a client and an agent is based on the completion of Tasks, which is an object defined by the protocol that has a lifecycle. The output of a task is called an Artefact;
Secure Collaboration: Agents can send each other messages to communicate context, replies, artefacts, or user instructions.
User experience negotiation: Each message includes “parts,” which is a fully formed piece of content, like a generated image. Each part has a specified content type, allowing client and remote agents to negotiate the correct format needed and explicitly include negotiations of the user’s UI capabilities–e.g., iframes, video, web forms, and more. Capability discovery: The protocol allows the Agents to advertise their capabilities, helping the clients identify which agent is best for performing a task; Capability discovery: Task and state management: The interaction between a client and an agent is based on the completion of Tasks, which is an object defined by the protocol that has a lifecycle. The output of a task is called an Artefact; Task and state management: Tasks, Artefact Secure Collaboration: Agents can send each other messages to communicate context, replies, artefacts, or user instructions. Secure Collaboration: User experience negotiation: Each message includes “parts,” which is a fully formed piece of content, like a generated image. Each part has a specified content type, allowing client and remote agents to negotiate the correct format needed and explicitly include negotiations of the user’s UI capabilities–e.g., iframes, video, web forms, and more. User experience negotiation: The most intriguing aspects are the Capability Discovery and User Experience Negotiation features, as they facilitate the establishment of Agent Marketplaces. Within these Marketplaces, suppliers can publish Agents, and clients can select the most suitable Agent to perform specific tasks. While this concept is undoubtedly promising and may be essential for the growth of the AI Agents market, it will require much more than the definition of a protocol for interaction to achieve this goal. Agent2Agent Protocol Concepts The protocol is based on a number of concepts, some of which are already familiar to those developing AI Agents: Agent Card: A public metadata file, describing the agent's capabilities, skills, endpoint URL, and authentication requirements, to be used in the discovery phase for selecting the Agent and understanding how to interact with it;
Server: An agent that implements the A2A protocol methods, as defined in the JSON specification;
Client: An application or another agent that consumes A2A services;
Task: The main unit of work for the Agent. It is initiated by the client and performed by the server, going through various states;
Message: Represents communication turns between the client and the agent. Each Message has a role and is composed of Parts;
Part: The fundamental content unit within a Message or an Artefact. A part can either be a text, a file or structured data;
Artefact: Represents outputs generated by the agent during the accomplishment of a Task. Artefacts, just like Messages, contain Parts.
Streaming: The protocol offers support for streaming, so that the server can use it to update the client about the state of long-running tasks. Agent Card: A public metadata file, describing the agent's capabilities, skills, endpoint URL, and authentication requirements, to be used in the discovery phase for selecting the Agent and understanding how to interact with it; Agent Card Server: An agent that implements the A2A protocol methods, as defined in the JSON specification; Server JSON specification JSON specification Client: An application or another agent that consumes A2A services; Client Task: The main unit of work for the Agent. It is initiated by the client and performed by the server, going through various states; Task states Message: Represents communication turns between the client and the agent. Each Message has a role and is composed of Parts; Message role Parts Part: The fundamental content unit within a Message or an Artefact. A part can either be a text, a file or structured data; Part Artefact: Represents outputs generated by the agent during the accomplishment of a Task. Artefacts, just like Messages, contain Parts. Artefact Streaming: The protocol offers support for streaming, so that the server can use it to update the client about the state of long-running tasks. Streaming State of the Art of the Agent2Agent Project A2A has just been announced to the public, and its specifications are now available on GitHub. At the present time, there is no official roadmap or production-ready implementation of the protocol, although Google is collaborating with partners to launch a production-ready version later this year (2025). The A2A GitHub repository contains a number of code samples in both TypeScript and Python, as well as a fairly comprehensive demo application. This application demonstrates the interaction between agents developed with different Agent Development Kits (ADK), as illustrated in the image below. A2A GitHub repository A2A GitHub repository code samples code samples demo application demo application Agent Development Kits (ADK) Agent Development Kits (ADK) Architecture of the A2A demo application (GitHub) GitHub GitHub This is enough to get started and experiment with the protocol, but before it is adopted in mission-critical projects, A2A needs to be integrated into the ecosystem of frameworks and tools built for adopting Agentic Workflows. Given the support of a large number of big names (interestingly, none of the companies that provide foundation models are present) working with Google on the protocol definition, there is little doubt that the tools will be here soon and that A2A will be integrated into the leading agent frameworks. List of the partners contributing to the Agent2Agent protocol (Google) Google Google Will A2A Replace Model Context Protocol (MCP)? Model Context Protocol (MCP) is a protocol designed by Anthropic that enables applications to provide context to Large Language Models. Anthropic describes MCP as the “USB-C port for AI applications”, providing a standardised way to connect LLMs to data sources and tools in the same fashion that USB allows connecting heterogeneous peripherals to devices. Model Context Protocol (MCP) Model Context Protocol (MCP) As Google explains, the A2A protocol is not meant to replace MCP. In fact, there is no overlap between MCP and A2A: they solve different problems and work at different levels of abstraction. A2A is designed to allow Agents to interact with other Agents, and MCP is designed to connect Large Language Models to tools, which in turn, connect them to services and data, as shown in the following image. Google explains Google explains Conclusions Agentic workflows have the potential to be a very interesting tool for solving complex problems, and open protocols such as A2A and MCP are certainly key enablers for the adoption of this technology. A2A is intended to become the protocol of choice for the interaction among agents and could be the basis for the development of marketplaces where agents can be advertised and made available to users. Large-scale adoption of Agentic workflows in enterprise-grade, mission-critical applications necessitates a multivendor ecosystem of tools and frameworks for the development, deployment, monitoring and tracing of multi-agent workflows. There are clear signals that the industry is moving in this direction, and A2A and MCP are an important part of this revolution. Links What are Agentic Workflows?;
Announcing the Agent2Agent Protocol (A2A);
A2A Documentation on GitHub;
A2A examples and demos;
A2A json specification;
Model Context Protocol documentation; What are Agentic Workflows?; What are Agentic Workflows? What are Agentic Workflows? Announcing the Agent2Agent Protocol (A2A); Announcing the Agent2Agent Protocol (A2A) Announcing the Agent2Agent Protocol (A2A) A2A Documentation on GitHub; A2A Documentation on GitHub A2A Documentation on GitHub A2A examples and demos; A2A examples and demos A2A examples and demos A2A json specification; A2A json specification A2A json specification Model Context Protocol documentation; Model Context Protocol documentation Model Context Protocol documentation If you found this article helpful, I’d love to connect with you on LinkedIn. If you found this article helpful, I’d love to connect with you on LinkedIn. LinkedIn LinkedIn

Getting to Know Google's Agent2Agent Protocol

Abhaya Uprety never imagined he would find himself living in a small studio apartment in New Mexico, carefully rationing groceries while building software for a company struggling to find its footing. An international student who graduated in May 2020, Uprety co-founded ParkStash—a startup initially focused on making parking reservations work in a way similar to short-term rental platforms. “We started off with the idea of being a sort of Airbnb for parking, focusing on everyday drivers,” Uprety said. “But as an international student fresh out of college, I didn’t have much room for trial and error. The stakes were always high.” His journey with ParkStash began by working alongside founder and CEO Sameer, an international student at San Jose State University. Inspired by his own struggles with limited campus parking, Sameer set out to build a parking platform. “He saw a problem and thought, ‘What if we connect people who have spare parking spaces with those who desperately need them?’” recalled Uprety. Over time, Uprety became the technology hub of the operation, single-handedly creating the company’s software stack. He handled mobile development using Flutter and Android, wrote server-side code in NodeJS, built dashboards with React, and more. By May 2020, when he officially joined ParkStash as a co-founder, the company was still focused on a business-to-consumer (B2C) model—allowing drivers to find and book private parking spots for a fee. Although people grasped the concept, the timing was challenging, especially given the pandemic’s impact on daily commuting patterns. Revenues remained inconsistent, and the burn rate was high. “I moved to New Mexico in early 2022,” said Uprety, “and lived in low-income housing, with no car, just enough to pay the rent and buy food. Still, I believed wholeheartedly in what we were doing.” From B2C to B2B: A Pivotal Shift By late 2022, ParkStash decided on a major pivot: from catering directly to individual drivers to providing a streamlined software solution for parking lot owners, property managers, and commercial sites. “We realized quickly that our real value was in helping the people who manage multiple parking spots,” remarked Uprety. “B2C was a fantastic idea in principle, but the market conditions and revenue model weren’t lining up for us. We needed a stable business structure, and that’s when we refocused our efforts on B2B.” The results of this pivot began to show almost immediately. ParkStash introduced subscription-based services for parking lot owners, offering tools like automated gate access control, permit management, camera-based monitoring, towing integrations, and a centralized dashboard for enforcement. “We demonstrate to property managers how their parking lot can turn into a revenue-producing asset,” said Uprety. “They see the potential to manage permits, catch violations, and even implement special promotions for residents or commercial tenants. That’s where the real growth started.” It did not take long for ParkStash to prove its value proposition. According to the INRIX 2017 Global Traffic Scorecard, American drivers spend an average of 17 hours per year searching for parking, costing them around 345 dollars each in wasted time, fuel, and emissions (INRIX 2017 Global Traffic Scorecard). Meanwhile, the International Parking & Mobility Institute has reported that up to 30 percent of city traffic can be caused by motorists circling for a spot (International Parking & Mobility Institute, 2019). ParkStash’s pivot to B2B tackled these issues by enabling large sites to optimize the use of parking space, reduce congestion, and offer reliable spots. “It’s not just about convenience,” Uprety explained. “We’re also addressing a real economic and environmental challenge.” By November 2022, the startup had completely embraced the B2B model. As of 2024, it has expanded to manage roughly 50 sites under contract, offering its clients a monthly subscription fee and optional add-ons like camera surveillance. “We are finally seeing real financial stability,” noted Uprety. “It’s been a whirlwind, but this pivot saved the company.” A Personal Triumph and New Industry Impact Part of ParkStash’s success story lies in Uprety’s resilience, honed by personal financial struggles and the uncertainties of running a startup. “There were times I doubted everything,” he said. “I was barely scraping by. But I’m passionate about programming, and I’m passionate about solving big problems. That passion kept me going.” His vision extends far beyond just booking parking spots. ParkStash’s B2B solution integrates with hardware installations, camera providers like Eagle Eye Networks, and towing services to form a cohesive ecosystem for property owners. “When we partner with a property, we handle as much as they need,” he explained. “Gate access, software dashboards, user-friendly permit assignment, a towing company for immediate violations, even a portal where folks can pay citations securely. We piece it all together so it’s one smooth experience.” The impact of this model goes beyond individual driveways or commercial lots. Properly managed parking means fewer cars circling the block, which can help lessen congestion and reduce carbon emissions. “It’s rewarding to think that we’re contributing to a more sustainable urban environment,” Uprety said. “Any time a driver reserves a spot and goes straight there without driving around in circles, that’s fuel saved, money saved, and time saved.” ParkStash’s evolution has also fueled new partnerships and job creation. By contracting with hardware installers, surveillance camera vendors, and towing operators, the startup is boosting smaller businesses. “One of my favorite parts is seeing how local entrepreneurs and technicians benefit from working with us,” Uprety added. “Our model depends on collaboration, so when we grow, they grow too.” For property owners, ParkStash’s platform can transform a previously overlooked or under-utilized asset—the parking lot—into a managed, revenue-generating space. “Parking can be complicated,” said Uprety. “But once we set up proper signage, digital enforcement, and streamlined user apps, these lots become smarter, more secure, and more profitable.” The company’s approach aligns with broader shifts toward smart cities, where digital infrastructure meets everyday needs like transportation and energy usage. “We’re not just building an app,” he stressed. “We’re helping shape what urban life could look like in the future.” Looking Ahead: Freedom, Innovation, and the Next Chapter For all of ParkStash’s accomplishments, Uprety remains keenly focused on personal and professional growth. Living on a shoestring budget in New Mexico until late 2023 taught him the value of efficiency—not just in code, but in day-to-day life. “We run lean on purpose,” he said. “We don’t do endless meetings. We talk to customers, gather feedback, and then get back to work. The best ideas come directly from the people who are actually using the product.” This customer-driven approach continues to shape ParkStash’s development roadmap. One key request from clients has been the ability to use mobile devices for license plate scanning, enabling easier on-site enforcement without costly specialized hardware. “We’re prototyping a feature that turns a phone or tablet camera into a plate reader,” Uprety revealed. “If an apartment manager can walk around the lot with their device and immediately see whether a car has a valid permit, that’s an instant productivity boost.” AI enhancements are also on the horizon. ParkStash aims to introduce voice-assisted technology that allows drivers to book a spot or extend a reservation without tapping through multiple screens. “I love exploring how AI can make our tools even more convenient,” Uprety said. “But we want to do it in a way that’s genuinely helpful, not just flashy. If it doesn’t solve a real user need, we’re not interested in building it.” Uprety’s personal satisfaction also stems from the company’s flexible culture—one that prioritizes hands-on coding and minimal bureaucracy. “I enjoy the freedom to innovate,” he noted. “We’re not weighed down by massive organizational hierarchies. Every feature we release is a direct result of real feedback and quick decision-making. That’s important to me because it feels honest, efficient, and human.” His journey from a financially strained international student to a proven thought leader in the parking tech space is a testament to resilience, passion, and strategic pivoting. “I look back on those months in New Mexico and think, if I hadn’t struggled, I wouldn’t have learned half of what I know now,” he said. “Sometimes, you have to run into walls to figure out how to build a better door.” As for ParkStash, the mission continues to expand. “We started with a simple concept: help people find a parking spot. Now, we’re helping businesses, communities, and entire cities manage a critical part of their infrastructure,” Uprety explained. “When I see a property manager rave about how easy it is to run their lots now, that’s the best possible validation.” Long-term, the company plans to keep improving its products while venturing into related areas of urban mobility. Its rising reputation, especially in the U.S. market, suggests potential partnerships with larger tech players and municipalities keen on adopting smart city solutions. “We’re building something that can evolve,” he said. “New features, new integrations, new ways to streamline how people move through cities. That’s a challenge that excites me.” In the end, what sets Uprety apart is the combination of technical prowess and a humble but determined spirit. By daring to pivot, ParkStash has charted a path that delivers meaningful solutions to property owners, drivers, and entire communities. “We’re just getting started,” he said, smiling. “The future of parking—and smart cities in general—holds endless possibilities. I’m just glad I get to help shape it, one line of code at a time. This story was authored under HackerNoon’s Business Blogging Program. This story was authored under HackerNoon’s Business Blogging Program. This story was authored under HackerNoon’s Business Blogging Program. This story was authored under HackerNoon’s Business Blogging Program. Business Blogging Program

Abhaya Uprety is Reshaping Parking One Line of Code at a Time

Abhaya Uprety يغير شكل الحجز في خط واحد من الكود في وقت واحد

"Abhaya Uprety" bir dəfə bir sıra kodu park edir

Abhaya Uprety wird Parken eine Zeile des Codes auf einmal neu gestalten

Η Abhaya Uprety επαναπροσδιορίζει τη στάθμευση μιας γραμμής κώδικα ταυτόχρονα

Abhaya Uprety está remodelando estacionamiento de una línea de código a la vez

Abhaya Uprety एक बार में एक पंक्ति कोड पार्किंग को फिर से आकार दे रहा है

Abhaya Upretyは、一度にコードの1行を駐車することを再構成しています。

Abhaya Uprety განკუთვნილია პარკირების ერთი ხაზი კოდი ერთხელ

Abhaya Uprety គឺជាការ re-shaping កន្លែងចែកចាយមួយបន្ទាត់នៃកូដនៅពេលមួយ

Abhaya Uprety нь нэг удаагийн код нь паркийг өөрчилж байна

Abhaya Uprety rishfaq parkimin e një rreshti të kodit në një kohë

Abhaya Uprety正在一次重塑停车场一行代码

Authors:
(1) Corby Rosset, Microsoft Research and Correspondence to corbyrosset@microsoft.com;
(2) Ching-An Cheng, Microsoft Research;
(3) Arindam Mitra, Microsoft Research;
(4) Michael Santacroce, Microsoft Research;
(5) Ahmed Awadallah, Microsoft Research and Correspondence to hassanam@microsoft.com;
(6) Tengyang Xie, Microsoft Research and Correspondence to tengyangxie@microsoft.com. Table of Links Abstract and 1 Introduction 2 Preliminaries 2.1 RLHF Based on Reward Models 2.2 RLHF with General Preferences 3 Direct Nash Optimization and 3.1 Derivation of Algorithm 1 3.2 Theoretical Analysis 4 Practical Algorithm – Iterative Contrastive Self-Improvement 5 Experiments and 5.1 Experimental Setup 5.2 Results and Analysis 6 Related Work 7 Conclusion and References Appendix A Extension to Regularized Preferences B Detailed Proofs C Additional Experimental Details 4 Practical Algorithm – Iterative Contrastive Self-Improvement In this section, we shift our focus to the algorithmic design of the practically scalable version of DNO, following the principles discussed in the last section. A primary challenge encountered in the implementation of the conceptual algorithm DNO (Algorithm 1) stems from the necessity to compute the expectation with respect to the preference function P under the current policy πt. Perhaps surprisingly, as we will show, all we need is a properly implemented iterative DPO-like contrastive learning algorithm. We present our the practical implementation of DNO in Algorithm 2 (DNO-Prct), which is a batched on-policy algorithm that conducts self-improvement iteratively via contrastive learning. One key consideration in our algorithmic design is that we only need to implicitly use the reward function rt. This comes from the specifically designed on-policy sampling, data filtering, and pair construction. While these specific design choices make DNO-Prct seem similar to simply performing DPO iteratively, there are significant reasons for these design decisions, as we will discuss below. Relationship between DNO-Prct and DPO. The reader may discern that DNO-Prct (Algorithm 2)—the practical implementation of DNO—can be described as an iterative version of the DPO algorithm. Such similarity is by design, intended to harness the simplicity and effectiveness of DPO (Rafailov et al., 2023) and build on empirical advancements from recent work that applies DPO iteratively (e.g., Yuan et al., 2024; Tran et al., 2024). Our experiments point to the importance of several design choices which help accommodate the general preferences, such as rankings derived from pair-wise win rates. More interestingly, our findings point to a surprising connection—that “a meticulously designed iterative DPO algorithm” could approach the Nash equilibrium of any given general preferences. Our general algorithmic framework—DNO (Algorithm 1)—is broader and fundamentally different from iterative DPO. For example, the DNO framework could also be directly extended to the regularized preference case (as discussed in Appendix A) or equipped with other advanced sample techniques (e.g., Liu et al., 2024b, RSO) as suggested by Theorem 1 for sample efficiency. On the other hand, although the soft policy iteration (or the KL-regularized reward optimization) is used in both DNO and DPO, they arise from fundamentally different reasons. For DNO, KL-regularization originates from online learning, no-regret learning through mirror descent (Nemirovskĳ and Yudin, 1983) or followthe-regularized-leader (FTRL) (Kalai and Vempala, 2005; Cesa-Bianchi and Lugosi, 2006; Shalev-Shwartz et al., 2012; Hazan et al., 2016). For DPO and PPO, the KL-regularization is an approximation for the total variation penalty to ensure monotonic improvement of the policy (Kakade and Langford, 2002; Schulman et al., 2015). Later, this approach was simplified by Schulman et al. (2017, PPO), and recently used for post-training LLMs (Ouyang et al., 2022). This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Corby Rosset, Microsoft Research and Correspondence to corbyrosset@microsoft.com; (2) Ching-An Cheng, Microsoft Research; (3) Arindam Mitra, Microsoft Research; (4) Michael Santacroce, Microsoft Research; (5) Ahmed Awadallah, Microsoft Research and Correspondence to hassanam@microsoft.com; (6) Tengyang Xie, Microsoft Research and Correspondence to tengyangxie@microsoft.com. Authors: Authors: (1) Corby Rosset, Microsoft Research and Correspondence to corbyrosset@microsoft.com; (2) Ching-An Cheng, Microsoft Research; (3) Arindam Mitra, Microsoft Research; (4) Michael Santacroce, Microsoft Research; (5) Ahmed Awadallah, Microsoft Research and Correspondence to hassanam@microsoft.com; (6) Tengyang Xie, Microsoft Research and Correspondence to tengyangxie@microsoft.com. Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 2 Preliminaries 2 Preliminaries 2.1 RLHF Based on Reward Models 2.1 RLHF Based on Reward Models 2.2 RLHF with General Preferences 2.2 RLHF with General Preferences 3 Direct Nash Optimization and 3.1 Derivation of Algorithm 1 3 Direct Nash Optimization and 3.1 Derivation of Algorithm 1 3.2 Theoretical Analysis 3.2 Theoretical Analysis 4 Practical Algorithm – Iterative Contrastive Self-Improvement 4 Practical Algorithm – Iterative Contrastive Self-Improvement 5 Experiments and 5.1 Experimental Setup 5 Experiments and 5.1 Experimental Setup 5.2 Results and Analysis 5.2 Results and Analysis 6 Related Work 6 Related Work 7 Conclusion and References 7 Conclusion and References Appendix Appendix A Extension to Regularized Preferences A Extension to Regularized Preferences B Detailed Proofs B Detailed Proofs C Additional Experimental Details C Additional Experimental Details 4 Practical Algorithm – Iterative Contrastive Self-Improvement In this section, we shift our focus to the algorithmic design of the practically scalable version of DNO, following the principles discussed in the last section. A primary challenge encountered in the implementation of the conceptual algorithm DNO (Algorithm 1) stems from the necessity to compute the expectation with respect to the preference function P under the current policy πt. Perhaps surprisingly, as we will show, all we need is a properly implemented iterative DPO-like contrastive learning algorithm. We present our the practical implementation of DNO in Algorithm 2 (DNO-Prct), which is a batched on-policy algorithm that conducts self-improvement iteratively via contrastive learning. One key consideration in our algorithmic design is that we only need to implicitly use the reward function rt. This comes from the specifically designed on-policy sampling, data filtering, and pair construction. While these specific design choices make DNO-Prct seem similar to simply performing DPO iteratively, there are significant reasons for these design decisions, as we will discuss below. Relationship between DNO-Prct and DPO. The reader may discern that DNO-Prct (Algorithm 2)—the practical implementation of DNO—can be described as an iterative version of the DPO algorithm. Such similarity is by design, intended to harness the simplicity and effectiveness of DPO (Rafailov et al., 2023) and build on empirical advancements from recent work that applies DPO iteratively (e.g., Yuan et al., 2024; Tran et al., 2024). Our experiments point to the importance of several design choices which help accommodate the general preferences, such as rankings derived from pair-wise win rates. More interestingly, our findings point to a surprising connection—that “a meticulously designed iterative DPO algorithm” could approach the Nash equilibrium of any given general preferences. Relationship between and “a meticulously designed iterative “a meticulously designed iterative algorithm” could approach the Nash equilibrium of any given general preferences. algorithm” could approach the Nash equilibrium of any given general preferences. Our general algorithmic framework—DNO (Algorithm 1)—is broader and fundamentally different from iterative DPO. For example, the DNO framework could also be directly extended to the regularized preference case (as discussed in Appendix A) or equipped with other advanced sample techniques (e.g., Liu et al., 2024b, RSO) as suggested by Theorem 1 for sample efficiency. On the other hand, although the soft policy iteration (or the KL-regularized reward optimization) is used in both DNO and DPO, they arise from fundamentally different reasons. For DNO, KL-regularization originates from online learning, no-regret learning through mirror descent (Nemirovskĳ and Yudin, 1983) or followthe-regularized-leader (FTRL) (Kalai and Vempala, 2005; Cesa-Bianchi and Lugosi, 2006; Shalev-Shwartz et al., 2012; Hazan et al., 2016). For DPO and PPO, the KL-regularization is an approximation for the total variation penalty to ensure monotonic improvement of the policy (Kakade and Langford, 2002; Schulman et al., 2015). Later, this approach was simplified by Schulman et al. (2017, PPO), and recently used for post-training LLMs (Ouyang et al., 2022). This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

How Contrastive Learning Helps AI Self-Improve

Hoe kontrastiewe leer help AI selfverbeter

Com l'aprenentatge contrastant ajuda a la IA a millorar-se

Πώς η αντίθετη μάθηση βοηθά την αυτοβελτίωση

Cómo el aprendizaje contrastante ayuda a la IA a mejorarse

Como o aprendizaxe contrastante axuda a AI a mellorarse

対照的な学習がAIが自らを改善するのを助ける方法

Cum învățarea contrastantă ajută AI să se îmbunătățească

Как контрастное обучение помогает ИИ самосовершенствовать

Fine-Tuning ChatGPT for Improved Performance in Bioinformatics

Authors:
(1) Jinge Wang, Department of Microbiology, Immunology & Cell Biology, West Virginia University, Morgantown, WV 26506, USA;
(2) Zien Cheng, Department of Microbiology, Immunology & Cell Biology, West Virginia University, Morgantown, WV 26506, USA;
(3) Qiuming Yao, School of Computing, University of Nebraska-Lincoln, Lincoln, NE 68588, USA;
(4) Li Liu, College of Health Solutions, Arizona State University, Phoenix, AZ 85004, USA and Biodesign Institute, Arizona State University, Tempe, AZ 85281, USA;
(5) Dong Xu, Department of Electrical Engineer and Computer Science, Christopher S. Bond Life Sciences Center, University of Missouri, Columbia, MO 65211, USA;
(6) Gangqing Hu, Department of Microbiology, Immunology & Cell Biology, West Virginia University, Morgantown, WV 26506, USA (Michael.hu@hsc.wvu.edu). Authors: Authors: (1) Jinge Wang, Department of Microbiology, Immunology & Cell Biology, West Virginia University, Morgantown, WV 26506, USA; (2) Zien Cheng, Department of Microbiology, Immunology & Cell Biology, West Virginia University, Morgantown, WV 26506, USA; (3) Qiuming Yao, School of Computing, University of Nebraska-Lincoln, Lincoln, NE 68588, USA; (4) Li Liu, College of Health Solutions, Arizona State University, Phoenix, AZ 85004, USA and Biodesign Institute, Arizona State University, Tempe, AZ 85281, USA; (5) Dong Xu, Department of Electrical Engineer and Computer Science, Christopher S. Bond Life Sciences Center, University of Missouri, Columbia, MO 65211, USA; (6) Gangqing Hu, Department of Microbiology, Immunology & Cell Biology, West Virginia University, Morgantown, WV 26506, USA (Michael.hu@hsc.wvu.edu). Table of Links Abstract and 1. Introduction Abstract and 1. Introduction 2. Omics 2. Omics 3. Genetics 3. Genetics 4. Biomedical Text Mining and 4.1. Performance Assessments across typical tasks 4. Biomedical Text Mining and 4.1. Performance Assessments across typical tasks 4.2. Biological pathway mining 4.2. Biological pathway mining 5. Drug Discovery 5. Drug Discovery 5.1. Human-in-the-Loop and 5.2. In-context Learning 5.1. Human-in-the-Loop and 5.2. In-context Learning 5.2 Instruction Finetuning 5.2 Instruction Finetuning 6. Biomedical Image Understanding 6. Biomedical Image Understanding 7. Bioinformatics Programming 7. Bioinformatics Programming 7.1 Application in Applied Bioinformatics 7.1 Application in Applied Bioinformatics 7.2. Biomedical Database Access 7.2. Biomedical Database Access 7.2. Online tools for Coding with ChatGPT 7.2. Online tools for Coding with ChatGPT 7.4 Benchmarks for Bioinformatics Coding 7.4 Benchmarks for Bioinformatics Coding 8. Chatbots in Bioinformatics Education 8. Chatbots in Bioinformatics Education 9. Discussion and Future Perspectives 9. Discussion and Future Perspectives Author Contributions, Acknowledgements, Conflict of Interest Statement, Ethics Statement, and References Author Contributions, Acknowledgements, Conflict of Interest Statement, Ethics Statement, and References 8. CHATBOTS IN BIOINFORMATICS EDUCATION The potential of integrating LLMs into bioinformatics education has attracted significant discussions. ChatGPT-3.5 achieves impressive performance in addressing Python programming exercises in an entrylevel bioinformatics course[12]. Beyond mere code generation, the utility of chatbots extends to proposing analysis plans, enhancing code readability, elucidating error messages, and facilitating language translation in coding tasks[105]. The effectiveness of a chatbot's response depends on the precision of human instructions, or prompts. In this context, Shue et al.[10] introduced the OPTIMAL model, which refines prompts through iterative interactions with a chatbot, mirroring the learning curve of bioinformatics novices assisted by such technologies. To navigate this evolving educational landscape, it becomes imperative to establish guidelines that enable students to critically assess outcomes and articulate constructive feedback to the chatbot for code improvement. Error messages, as one form of such feedback, turn out to be an effective way to boost the coding efficiency of ChatGPT across various studies[10, 11, 96]. The convenience of using chatbots for coding exercises poses a risk of fostering AI overreliance, leading to a superficial understanding of the underlying concepts[12, 13, 106]. This AI reliance could undermine students' performance in summative assessments[12]. Innovative evaluation strategies, such as generating multiple-choice questions from student-submitted code to gauge their understanding[107], are needed to counteract this challenge. Such methodologies should aim to deepen students' grasp of the material, ensuring their in-depth understanding of coding concepts. The art of crafting effective prompts emerges as a critical skill that complements traditional programming competencies. Feedback from a pilot study involving graduate students interacting with ChatGPT for coding highlights the challenges in generating impactful prompts[108]. This prompt-related psychological strain may discourage students from using the chatbot[13]. Instructors need to equip students with prompting skills known to enhance coding efficacy, such as one/few-shot by providing examples and stimulus prompting by listing function/package names[13]. In addition, the development of a repository featuring carefully crafted prompts, accompanied by quality metrics, reference code, and outcomes, could serve as a valuable resource for students to learn coding through prompting[10, 13]. In conclusion, while chatbots demonstrate potential as educational tools, their efficacy and effectiveness have not yet been systematically evaluated in classroom settings with controlled experiments. The use of chatbots should be viewed as supplementary to traditional education methodologies[10, 12, 13]. Meanwhile, new assessment methodologies are needed to measure the pedagogical value of chatbots in enhancing bioinformatics learning without diminishing the depth of understanding of concepts and analytical skills. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Chatbots in Bioinformatics Education

I know the word layoff is on the mind of a majority of corporate workers right now, and it is no joke. This job market is brutal, and the number of texts I get from talented, highly-skilled people unable to find a job is revolting. We are told we must adjust to a new reality where AI would steal most jobs, leaving us with “nothing” and having no economic prospects to sustain our lives. But is this fear really about AI? Or is it about how short-lived our certainties have always been? AI may ironically free us from the very system we fear losing. AI may ironically free us from the very system we fear losing. There are many things that become so ominous to our life that we don’t realize how short a time they have been around. Landline phones, televisions, Facebook — all came into our lives as disruptive technologies and stopped being the center of our lives before we realized it. The same is going to happen to the conventional 9-to-5 workday. While so deeply ingrained in our modern consciousness, it represents a surprisingly brief experiment in the long arc of human labor history. 9-to-5 workday a surprisingly brief experiment in the long arc of human labor history. As AI reshapes how things get done, we stand to witness another transformation — one that is seeing us return to more goal-oriented rather than time-oriented work patterns. The Final Nail In the Coffin Of Industrial Revolution Labor Patterns A world of year-end reviews, clocking 40 hours a week, and more corporate shenanigans that people who have held real jobs, unlike me, tell me about have brought comfort to people they are realizing is being pulled from under them. Yet, this wasn’t the case for like 95% of human history. We organized our labor around immediate needs and natural cycles. The strict delineation of work time versus personal time would have seemed peculiar to our ancestors until the 18th century. You all think a 4-day work week is a new progressive, modern way of working that we are experimenting with. Blah! Hunter-gatherers in the Stone Age worked just three to five hours daily. The radical transformation of work schedules came with the Industrial Revolution, when labor became increasingly detached from natural rhythms and human needs. By 1817, people were working unprecedented 80-100 hour weeks. Factory workers in mid-19th century England endured punishing 16-hour days, 311 days annually. Workers lost control, prompting organized resistance to excessive work hours. The breakthrough in the struggle came in the early 20th century. In 1926, Henry Ford implemented the 40-hour workweek after discovering diminishing returns from longer hours. In 1940, the 40-hour workweek became U.S. law, establishing what we now know as the standard schedule. The term "9-to-5" became so emblematic of standardized work schedules that it entered popular culture. In historical context, this rigid time-based approach to work has existed for less than a century — a mere blip in humanity's 300,000-year history. I think, substantial progress in AI would put the final nail in the coffin of everything that was wrong with the Industrial Revolution and the exploitation of labor that came with it. While we put machines to work as manufacturing takes a more important role than ever before. The Bigger Picture of AI-Driven Layoffs Corporations exist for a single reason — to fetch returns for shareholders. It doesn’t matter how many times your manager says you’re like a family. So, there are two things that are set to pan out. Either a company optimizes its workflow, doing more with a smaller headcount, or sees its business collapse. In the next decade, it may become a more financially safe decision to be an entrepreneur or independent creator than being a 9-to-5er. In the next decade, it may become a more financially safe decision to be an entrepreneur or independent creator than being a 9-to-5er. The jobs that remain relevant in the AI era will likely emphasize what humans accomplish rather than simply how long they work — a change that, viewed historically, represents less a revolution than a homecoming to more natural patterns of human productivity. less a revolution than a homecoming to more natural patterns of human productivity. Estimated work hours falling post-AGI, continuing the trend from the peak of the early industrial age in the 18th century. A recent Pew study found that 52% of Americans are worried about how AI is used in the workplace, yet only 16% incorporate it in any way in their work. I think there lies the crux of the issue. If you’re more worried about keeping your job than swinging the hammer, you’ll miss the nail entirely, in my humble view. If you think this is some hypothetical novelty of the future — just look at the past couple of years. Companies like Cursor, Midjourney, and ElvenLabs are launching path-breaking products with a team of 10 people or so. Meanwhile, remind me, what is IBM doing? AI is going to take down more inefficiently-run large corporations than it does talented people. AI is going to take down more inefficiently-run large corporations than it does talented people. It is going to play a role in balancing out the scales in power in favor of the passionate, hard-working, obsessed little guy; take my word for it. This article was published first as part of the Artificially Boosted newsletter. This article was published first as part of the Artificially Boosted newsletter. Artificially Boosted

If Getting Laid-off is Bad, Wait Till You See What AI Does to Corporates

Disclaimer: Parts of this article are purely fictional. ZBlockBuilders Inc., its CEO, and all accompanying scenarios are made up to help explain complex blockchain concepts in a fun, relatable way. Any resemblance to real companies or people is purely coincidental. Disclaimer: Parts of this article are purely fictional. ZBlockBuilders Inc., its CEO, and all accompanying scenarios are made up to help explain complex blockchain concepts in a fun, relatable way. Any resemblance to real companies or people is purely coincidental. Disclaimer: Parts of this article are purely fictional. ZBlockBuilders Inc., its CEO, and all accompanying scenarios are made up to help explain complex blockchain concepts in a fun, relatable way. Any resemblance to real companies or people is purely coincidental. Don’t lose the farm. Every CEO and their team needs to understand how blockchain nodes work to keep their revenues and respect secure. Don’t lose the farm. Every CEO and their team needs to understand how blockchain nodes work to keep their revenues and respect secure. Don’t lose the farm. Every CEO and their team needs to understand how blockchain nodes work to keep their revenues and respect secure. Congratulations, Reader. You have just been invited to join the team, as Chief Executive Officer, at ZBlockBuilders Inc., a Web3 technology company looking to extend financial coverage for 1.5 billion financially underserved people across the globe. There are several targets on your mind and, of course, a few concerns about how you’ll accomplish your company’s immediate objective of bringing these into the fold. In other words, you want to democratize the financial system by first educating your potential market.  You’ve been offered a fat paycheck and a tall order so you’ve hired me as your adviser to help with this task. Node is the Codeword Our first marketing campaign is to introduce the idea of nodes to our target market. I’ve come up with an interesting idea: let’s show them how dangerous the current fintech system could get. Bands of bad actors roam the internet, empowered by sophisticated hacking software, snooping on customers, hijacking entire banking apps, and carting off troves of customer data. We develop an ad that shows how we can prevent this cataclysmic conclusion by sharing that our technology depends on independent decentralized computers (known as nodes), all functioning as “checkers-and-balancers” on each other. No transaction is added to the blockchain unless and until all nodes approve that it is valid: a parliament of protocols if you like. So, if they’re out on a beach in Mexico chilling, half-drunk, with a Piña Colada in hand, they can be sure that no unauthorized transactions are added to the blockchain (or approved) until they give the word. This strikes you as really important because you recall that when you were CFO at LibertyWonk Inc., a Web-2 based fintech, some sleazy fellow with passable tech credentials capered away with $ 30 million through a Business Email Compromise scheme, ruining lives and LibertyWonk Inc. itself. You laugh, telling me about a certain Hushpuppi, who commandeered atleast $24 million from his mastery of BEC and other cybercrimes. a parliament of protocols a parliament of protocols Hushpuppi, Hushpuppi Copy That! Now, you’re ensconced in your aubergine-colored leather armchair grinning at me. You’re starting to like this. “Tell me more,” you say. I go on sharing that we’ll introduce a series of social media posts reiterating that each of our nodes stores a copy of the entire node system, allowing anybody to verify the provenance of every transaction. In other words, there’s no hiding behind forged records. I flash a web page from my mobile device in front of you, taking you back on memory lane to remind you of the lessons learned from the investigations of the Bank of Credit and Commerce International (BCCI). Investigations which commenced in 1992 and concluded in 2012, cost $656 million paid to firms of lawyers and accountants and resulted in innumerable heartbreaks in victims of the humongous heist. Bank of Credit and Commerce International (BCCI). Bank of Credit and Commerce International (BCCI) Calling all Computers! Nodes (computers, remember?) are linguistic linkers. In other words, they like to communicate with each other just as humans do. Beneath the humming sounds of their hardware, they are asking each other questions about the state of the blockchain just the same way you, as the CEO, would ring the CFO to make yet another inquiry about the financial health of ZBlockBuilders Inc., on a Friday afternoon as you’re getting set to proceed on vacation. While nodes may not go on holiday, they’re still super concerned about how well records are being kept, and believe me, they’ll ask. Something like “hey, are there any new transactions and blocks on our chain, and what are their details?” They constantly check on each other to see how things are going and this involves a process called the consensus mechanism. This simply means that they use a special mode of confirmation to check the state of the blockchain. I’ll share two types of consensus mechanisms. How do Consensus Mechanisms Function? Proof of Work This is native to the Bitcoin core and what it does is require a block miner to solve some truly complex piece of math to create a new block. Once the enterprising mathematician unlocks a new block, it immediately is broadcast to the entire blockchain (just like when Rafiki proudly held the newborn lion cub for all the jungle to see in The Lion King). It more or less says “A new block is born” but without the drama of lions and other animals involved. Proof of Stake Now, this is native to Ethereum and is somewhat different from how things play out in the Bitcoin core. With the proof of stake mechanism, you’re a validator and not a miner (note the distinction). So you’re not actively mining as such. Instead, you’re expected to stake 32 ETH (thirty-two Ethereum) and have a validator software ready and running in the hopes that you’re selected to propose new blocks to be added to the blockchain. If you act right, you’ll get some rewards for it. If you’re considered irresponsible (acting badly or staying offline for too long: slashing) you’re punished by having to lose some of your ETH (Ethereum). Decentralized and Defended The overwhelming beauty of nodes in the blockchain is that each node is a separate entity of its own, acting independently by itself but unanimously with its peers.  Nodes cannot be compelled to approve transactions or fooled by the witty disguises of questing cybercriminals with suggestive emails, until certain rules are collectively adhered to. There is no central system, as we have in current fintech configurations, that can be manipulated or harassed to relinquish value. Nodes function in a trustless environment. That does not mean that they abhor trust, rather they do not rely on trust or goodwill to function. They are ruthless and reliable executors. In Conclusion Essentially, nodes are the librarians, accountants and security guards of the blockchain. They communicate with each other to keep records of the particulars of the blockchain, ensuring that all transactions are properly accounted for and blocking (no pun intended) any unauthorized interactions. Lunch Break,Chief I see that we’re getting somewhere with my marketing campaign proposals because you’re handing me a glass of orange juice. I’m getting quite hungry too. So, if you like, I’ll come back to offer some more advice after the break. Featured Image courtesy of Lalmch on Pixabay Featured Image courtesy of Lalmch on Pixabay Lalmch on Pixabay

Stories, News & Mentions about AECOM

Categories

Trending Topics

Stories, News & Mentions about AECOM

Aecom’s Stories on HackerNoon

Aecom's News & Mentions

Categories

Trending Topics