The race to sign Alessandro Buongiorno has begun and AC Milan are joined by two of their domestic rivals ahead of the summer, according to a report.

CM: How Conte's call affects the Buongiorno sweepstakes as Milan continue talks

CONTE

conte

Table of Links Abstract and Introduction
SylloBio-NLI
Empirical Evaluation
Related Work
Conclusions
Limitations and References A. Formalization of the SylloBio-NLI Resource Generation Process B. Formalization of Tasks 1 and 2 C. Dictionary of gene and pathway membership D. Domain-specific pipeline for creating NL instances and E Accessing LLMs F. Experimental Details G. Evaluation Metrics H. Prompting LLMs - Zero-shot prompts I. Prompting LLMs - Few-shot prompts J. Results: Misaligned Instruction-Response K. Results: Ambiguous Impact of Distractors on Reasoning L. Results: Models Prioritize Contextual Knowledge Over Background Knowledge M Supplementary Figures and N Supplementary Tables L Results: Models Prioritize Contextual Knowledge Over Background Knowledge The lack of statistically significant differences (Fig. 7) in accuracy between biologically factual and artificial datasets across generalized modus ponens and generalized modus tollens schemes suggests that the models’ reasoning capabilities rely more on stated contextual knowledge and logical structure than on pre-existing background knowledge. This holds true for both accuracy and reasoning accuracy, as well as in both ZS and FS settings: models that perform well on a given scheme maintain their performance even when factual gene names are replaced by synthetic names, and the same consistency is observed for models with weaker performance. This ability to maintain accuracy with synthetic gene names in the artificial set demonstrates that models can abstract and apply logical reasoning independently of their internal domain-specific knowledge. Authors:
(1) Magdalena Wysocka, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom;
(2) Danilo S. Carvalho, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom and Department of Computer Science, Univ. of Manchester, United Kingdom;
(3) Oskar Wysocki, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom and ited Kingdom 3 I;
(4) Marco Valentino, Idiap Research Institute, Switzerland;
(5) André Freitas, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom, Department of Computer Science, Univ. of Manchester, United Kingdom and Idiap Research Institute, Switzerland. This paper is available on arxiv under CC BY-NC-SA 4.0 license. Table of Links Abstract and Introduction SylloBio-NLI Empirical Evaluation Related Work Conclusions Limitations and References Abstract and Introduction Abstract and Introduction SylloBio-NLI SylloBio-NLI Empirical Evaluation Empirical Evaluation Related Work Related Work Conclusions Conclusions Limitations and References Limitations and References A. Formalization of the SylloBio-NLI Resource Generation Process A. Formalization of the SylloBio-NLI Resource Generation Process B. Formalization of Tasks 1 and 2 B. Formalization of Tasks 1 and 2 C. Dictionary of gene and pathway membership C. Dictionary of gene and pathway membership D. Domain-specific pipeline for creating NL instances and E Accessing LLMs D. Domain-specific pipeline for creating NL instances and E Accessing LLMs F. Experimental Details F. Experimental Details G. Evaluation Metrics G. Evaluation Metrics H. Prompting LLMs - Zero-shot prompts H. Prompting LLMs - Zero-shot prompts I. Prompting LLMs - Few-shot prompts I. Prompting LLMs - Few-shot prompts J. Results: Misaligned Instruction-Response J. Results: Misaligned Instruction-Response K. Results: Ambiguous Impact of Distractors on Reasoning K. Results: Ambiguous Impact of Distractors on Reasoning L. Results: Models Prioritize Contextual Knowledge Over Background Knowledge L. Results: Models Prioritize Contextual Knowledge Over Background Knowledge M Supplementary Figures and N Supplementary Tables M Supplementary Figures and N Supplementary Tables L Results: Models Prioritize Contextual Knowledge Over Background Knowledge The lack of statistically significant differences (Fig. 7) in accuracy between biologically factual and artificial datasets across generalized modus ponens and generalized modus tollens schemes suggests that the models’ reasoning capabilities rely more on stated contextual knowledge and logical structure than on pre-existing background knowledge. This holds true for both accuracy and reasoning accuracy, as well as in both ZS and FS settings: models that perform well on a given scheme maintain their performance even when factual gene names are replaced by synthetic names, and the same consistency is observed for models with weaker performance. This ability to maintain accuracy with synthetic gene names in the artificial set demonstrates that models can abstract and apply logical reasoning independently of their internal domain-specific knowledge. Authors: (1) Magdalena Wysocka, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom; (2) Danilo S. Carvalho, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom and Department of Computer Science, Univ. of Manchester, United Kingdom; (3) Oskar Wysocki, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom and ited Kingdom 3 I; (4) Marco Valentino, Idiap Research Institute, Switzerland; (5) André Freitas, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom, Department of Computer Science, Univ. of Manchester, United Kingdom and Idiap Research Institute, Switzerland. Authors: Authors: (1) Magdalena Wysocka, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom; (2) Danilo S. Carvalho, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom and Department of Computer Science, Univ. of Manchester, United Kingdom; (3) Oskar Wysocki, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom and ited Kingdom 3 I; (4) Marco Valentino, Idiap Research Institute, Switzerland; (5) André Freitas, National Biomarker Centre, CRUK-MI, Univ. of Manchester, United Kingdom, Department of Computer Science, Univ. of Manchester, United Kingdom and Idiap Research Institute, Switzerland. This paper is available on arxiv under CC BY-NC-SA 4.0 license. This paper is available on arxiv under CC BY-NC-SA 4.0 license. available on arxiv

LLMs Rely on Contextual Knowledge Over Background Knowledge

Authors:
(1) Albert Gu, Machine Learning Department, Carnegie Mellon University and with equal contribution;
(2) Tri Dao, Department of Computer Science, Princeton University and with equal contribution. Table of Links Abstract and 1 Introduction 2 State Space Models 3 Selective State Space Models and 3.1 Motivation: Selection as a Means of Compression 3.2 Improving SSMs with Selection 3.3 Efficient Implementation of Selective SSMs 3.4 A Simplified SSM Architecture 3.5 Properties of Selection Mechanisms 3.6 Additional Model Details 4 Empirical Evaluation and 4.1 Synthetic Tasks 4.2 Language Modeling 4.3 DNA Modeling 4.4 Audio Modeling and Generation 4.5 Speed and Memory Benchmarks 4.6 Model Ablations 5 Discussion 6 Conclusion and References A Discussion: Selection Mechanism B Related Work C Mechanics of Selective SSMs D Hardware-aware Algorithm For Selective SSMs E Experimental Details and Additional Results 3.5 Properties of Selection Mechanisms The selection mechanism is a broader concept that can be applied in different ways, such as to more traditional RNNs or CNNs, to different parameters (e.g. A in Algorithm 2), or using different transformations s(x). 3.5.1 Connection to Gating Mechanisms We highlight the most important connection: the classical gating mechanism of RNNs is an instance of our selection mechanism for SSMs. We note that the connection between RNN gating and the discretization of continuous-time systems is well established (Funahashi and Nakamura 1993; Tallec and Ollivier 2018). In fact, Theorem 1 is an improvement of Gu, Johnson, Goel, et al. (2021, Lemma 3.1) generalizing to the ZOH discretization and input-dependent gates (proof in Appendix C). More broadly, ∆ in SSMs can be seen to play a generalized role of the RNN gating mechanism. In line with prior work, we adopt the view that discretization of SSMs is the principled foundation of heuristic gating mechanisms. 3.5.2 Interpretation of Selection Mechanisms We elaborate on two particular mechanistic effects of selection. Filtering Context. It has been empirically observed that many sequence models do not improve with longer context (F. Shi et al. 2023), despite the principle that more context should lead to strictly better performance. An explanation is that many sequence models cannot effectively ignore irrelevant context when necessary; an intuitive example are global convolutions (and general LTI models). On the other hand, selective models can simply reset their state at any time to remove extraneous history, and thus their performance in principle improves monotonicly with context length (e.g. Section 4.3.2). Additionally, we elaborate on effects of each selective parameter. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Albert Gu, Machine Learning Department, Carnegie Mellon University and with equal contribution; (2) Tri Dao, Department of Computer Science, Princeton University and with equal contribution. Authors: Authors: (1) Albert Gu, Machine Learning Department, Carnegie Mellon University and with equal contribution; (2) Tri Dao, Department of Computer Science, Princeton University and with equal contribution. Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 2 State Space Models 2 State Space Models 3 Selective State Space Models and 3.1 Motivation: Selection as a Means of Compression 3 Selective State Space Models and 3.1 Motivation: Selection as a Means of Compression 3.2 Improving SSMs with Selection 3.2 Improving SSMs with Selection 3.3 Efficient Implementation of Selective SSMs 3.3 Efficient Implementation of Selective SSMs 3.4 A Simplified SSM Architecture 3.4 A Simplified SSM Architecture 3.5 Properties of Selection Mechanisms 3.5 Properties of Selection Mechanisms 3.6 Additional Model Details 3.6 Additional Model Details 4 Empirical Evaluation and 4.1 Synthetic Tasks 4 Empirical Evaluation and 4.1 Synthetic Tasks 4.2 Language Modeling 4.2 Language Modeling 4.3 DNA Modeling 4.3 DNA Modeling 4.4 Audio Modeling and Generation 4.4 Audio Modeling and Generation 4.5 Speed and Memory Benchmarks 4.5 Speed and Memory Benchmarks 4.6 Model Ablations 4.6 Model Ablations 5 Discussion 5 Discussion 6 Conclusion and References 6 Conclusion and References A Discussion: Selection Mechanism A Discussion: Selection Mechanism B Related Work B Related Work C Mechanics of Selective SSMs C Mechanics of Selective SSMs D Hardware-aware Algorithm For Selective SSMs D Hardware-aware Algorithm For Selective SSMs E Experimental Details and Additional Results E Experimental Details and Additional Results 3.5 Properties of Selection Mechanisms The selection mechanism is a broader concept that can be applied in different ways, such as to more traditional RNNs or CNNs, to different parameters (e.g. A in Algorithm 2), or using different transformations s(x). 3.5.1 Connection to Gating Mechanisms 3.5.1 Connection to Gating Mechanisms We highlight the most important connection: the classical gating mechanism of RNNs is an instance of our selection mechanism for SSMs. We note that the connection between RNN gating and the discretization of continuous-time systems is well established (Funahashi and Nakamura 1993; Tallec and Ollivier 2018). In fact, Theorem 1 is an improvement of Gu, Johnson, Goel, et al. (2021, Lemma 3.1) generalizing to the ZOH discretization and input-dependent gates (proof in Appendix C). More broadly, ∆ in SSMs can be seen to play a generalized role of the RNN gating mechanism. In line with prior work, we adopt the view that discretization of SSMs is the principled foundation of heuristic gating mechanisms. 3.5.2 Interpretation of Selection Mechanisms 3.5.2 Interpretation of Selection Mechanisms We elaborate on two particular mechanistic effects of selection. Filtering Context. It has been empirically observed that many sequence models do not improve with longer context (F. Shi et al. 2023), despite the principle that more context should lead to strictly better performance. An explanation is that many sequence models cannot effectively ignore irrelevant context when necessary; an intuitive example are global convolutions (and general LTI models). On the other hand, selective models can simply reset their state at any time to remove extraneous history, and thus their performance in principle improves monotonicly with context length (e.g. Section 4.3.2). Filtering Context. Additionally, we elaborate on effects of each selective parameter. 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 Selection Mechanisms Address Long-Context Limitations in AI Sequence Models

Recently, I have been taking some sessions on the road to mastering Data Structures and Algorithms and Competitive Programming. Leetcode contests are perfect for those who want a grip on their DSA concepts and want to experience competitive pressure. Now, you might be thinking, “Do we really need to read this article and who’s this person?” So, I must tell you that I have been coding properly for like past 6–7 months only. I started Leetcode contests probably in like September of last year. Since then, I have given almost all the contests, and it took me just 15 contests to get into the top 3% of Leetcoders. You can do the same with a little bit of dedication and motivation. Just keep these points in mind: Don’t Miss Contests No matter if you are busy with your exams or day-to-day work, you must take some of your time and give it to the contest. Leetcode Biweekly contests are held alternating Saturdays at 8 pm IST, and Weekly contests are held every Sunday at 8 am IST. If by any chance you missed any contest, attempt a virtual contest. Virtual contests give you an idea of what your ranking would be, but it will not change your rating. Upsolve Problems Generally, people participate and then post contests that don’t try to solve unsolved problems or try to find a more optimized approach for the problem they have solved. Note this: if you do not upsolve, you can’t learn anything from the contest. Those who don’t learn from their mistakes are bound to repeat them. Like if you have solved A and B questions. Then give one day to review the solutions you have solved and one day to solve the C and D questions. Topics to Cover Contests on platforms like Leetcode, GeeksForGeeks(GFG), and CodeStudio are more based on DSA and not on competitive concepts. So for these contests, you should have a good grip on below mentioned topics: Arrays


Linked Lists


Stack/Queue/Priority Queue


Trees: Traversal questions and the DFS approach are important


Graphs: Graph algorithms like Dijkstra and Floyd Warshall are important.


Greedy Algorithm


Dynamic Programming


Union Find


Two Pointer Approach


Sliding Window Hold on; Trees, Graphs, and Dynamic Programming take time. These are a bit lengthy topics and need more practice. Refer to my curated repository for topic-wise codes for various problems and concepts. P.S: First try to solve the problem on your own. Follow: https://github.com/Kanav-Arora/DSA-Guide for DSA Roadmap and Solutions. Don’t Directly Jump to the Solution This is one of the most important points. Directly jumping to the solution makes you handicapped and will not improve your problem approachability. First, give it a thought, and try to dry-run your solutions on given test cases. Solve it on a paper and then submit it. Then only should you consider looking for the solution. Now, the most common mistake people make is they prefer text solutions over video solutions but text solutions can’t have that much of an understanding. You need to get the intuition of the problem so that you can solve it again on your own. Believe it or not, Leetcode problems are the same or somewhat similar to the interview problems. So, you must solve these if you are appearing for a tech interview. Thanks for reading! If you like the article, please give it a clap👏, and follow the publication for more such stories. Follow me on LinkedIn and GitHub. Recently, I have been taking some sessions on the road to mastering Data Structures and Algorithms and Competitive Programming. Leetcode contests are perfect for those who want a grip on their DSA concepts and want to experience competitive pressure. Now, you might be thinking, “Do we really need to read this article and who’s this person?” So, I must tell you that I have been coding properly for like past 6–7 months only. I started Leetcode contests probably in like September of last year. Since then, I have given almost all the contests, and it took me just 15 contests to get into the top 3% of Leetcoders. You can do the same with a little bit of dedication and motivation. Just keep these points in mind: Don’t Miss Contests Don’t Miss Contests No matter if you are busy with your exams or day-to-day work, you must take some of your time and give it to the contest. Leetcode Biweekly contests are held alternating Saturdays at 8 pm IST, and Weekly contests are held every Sunday at 8 am IST. If by any chance you missed any contest, attempt a virtual contest. Virtual contests give you an idea of what your ranking would be, but it will not change your rating. Upsolve Problems Upsolve Problems Generally, people participate and then post contests that don’t try to solve unsolved problems or try to find a more optimized approach for the problem they have solved. Note this: if you do not upsolve, you can’t learn anything from the contest. Those who don’t learn from their mistakes are bound to repeat them. Those who don’t learn from their mistakes are bound to repeat them. Those who don’t learn from their mistakes are bound to repeat them. Like if you have solved A and B questions. Then give one day to review the solutions you have solved and one day to solve the C and D questions. Topics to Cover Topics to Cover Contests on platforms like Leetcode, GeeksForGeeks(GFG), and CodeStudio are more based on DSA and not on competitive concepts. So for these contests, you should have a good grip on below mentioned topics: Arrays Linked Lists Stack/Queue/Priority Queue Trees: Traversal questions and the DFS approach are important Graphs: Graph algorithms like Dijkstra and Floyd Warshall are important. Greedy Algorithm Dynamic Programming Union Find Two Pointer Approach Sliding Window Arrays Arrays Linked Lists Linked Lists Stack/Queue/Priority Queue Stack/Queue/Priority Queue Trees: Traversal questions and the DFS approach are important Trees: Traversal questions and the DFS approach are important Graphs: Graph algorithms like Dijkstra and Floyd Warshall are important. Graphs: Graph algorithms like Dijkstra and Floyd Warshall are important. Greedy Algorithm Greedy Algorithm Dynamic Programming Dynamic Programming Union Find Union Find Two Pointer Approach Two Pointer Approach Sliding Window Sliding Window Hold on; Trees, Graphs, and Dynamic Programming take time. These are a bit lengthy topics and need more practice. Refer to my curated repository for topic-wise codes for various problems and concepts. P.S: First try to solve the problem on your own. Follow: https://github.com/Kanav-Arora/DSA-Guide for DSA Roadmap and Solutions. Follow: https://github.com/Kanav-Arora/DSA-Guide for DSA Roadmap and Solutions. https://github.com/Kanav-Arora/DSA-Guide Don’t Directly Jump to the Solution Don’t Directly Jump to the Solution This is one of the most important points. Directly jumping to the solution makes you handicapped and will not improve your problem approachability. First, give it a thought, and try to dry-run your solutions on given test cases. Solve it on a paper and then submit it. Then only should you consider looking for the solution. Now, the most common mistake people make is they prefer text solutions over video solutions but text solutions can’t have that much of an understanding. You need to get the intuition of the problem so that you can solve it again on your own. Believe it or not, Leetcode problems are the same or somewhat similar to the interview problems. So, you must solve these if you are appearing for a tech interview. Thanks for reading! If you like the article, please give it a clap👏, and follow the publication for more such stories. Follow me on LinkedIn and GitHub . LinkedIn LinkedIn GitHub

How to Go From Noob to Master: Leetcode Contests

Hey Hackers! Great news! If you haven’t had a chance to submit your entry for the AI Writing Contest, the submission deadline has been extended to December 16, 2024. You now have two extra weeks to journey through the world of data collection for AI and LLM training for a shot at $2,500 in prizes. Enter Now!
Choose a writing template from the list below, fill it out, and submit it.

The Future of Data Collection for AI and LLMs
Using Bright Data for AI and LLM Training
Web Scraping for AI and LLM Training Data AI is everywhere—from kids using it at school to professionals relying on it in their work, and even in your mom's Facebook group chat generating stickers. This contest shifts the focus from the flashy front end most users are familiar with to the unsung hero: data. Since we launched in September, participants have shared unique insights on data collection for AI and LLM training. For example, @ayinketh discusses overcoming traditional web scraping challenges with Bright Data in his entry, while @adnanaleeza uncovers a sentiment analysis hack in hers. Others explored everything from AI hallucinations to industry analysis. With just two weeks left and $2,500 still up for grabs, now’s the perfect time to enter the contest. Tips for Entering the AI Writing Contest 1. Learn About Bright Data Start by watching the Introduction to Bright Datasets video below to understand how Bright Data works. https://youtu.be/sUDROEVLqEk?feature=shared&embedable=true 2. Explore Bright Data Take advantage of Bright Data’s $15 credit to explore its  Dataset Marketplace and Web Data for AI. It’s a great way to get hands-on experience with the tools you’ll be using. 3. Fill Out a Contest Template Choose a template from the list below, fill it out, and submit your entry. The Future of Data Collection for AI and LLMs
Using Bright Data for AI and LLM Training
Web Scraping for AI and LLM Training Data That’s all! By following these steps, you could win any of the prizes below: Prize Breakdown for the AI Writing Contest

$$$



Best Bright Data Story

$1000



Best AI Story

$1000



Best Web Scraping Story

$500 Rules & Guidelines Must be 18+ to enter (no location restrictions)
Create a HackerNoon account
You can enter the contest by adding the #ai tag to your story.
No AI-generated entries. (Read: the acceptable way to use AI on HackerNoon) FAQs Can I Write Under a Pen Name? Yes!You can use your real name, or a pseudonym when setting up your HackerNoon profile. How Long Will the Contest Run? The contest is open until December 16, 2024. How are the winners selected? After the contest ends, we’ll review submitted entries and shortlist the stories that receive the most eyeballs (real humans, not bots!).
Next, the shortlisted stories will be voted on by HackerNoon staff.
Then, the best in each prize category; AI, Bright Data, and Web Scraping, will be selected and announced Ready to enter the AI writing contest? Start a draft now, or enter the contest using any of these templates:

Using Bright Data for AI and LLM Training
The Future of Data Collection for AI and LLMs
Web Scraping for AI and LLM Data Collection Good luck! Hey Hackers! Great news! If you haven’t had a chance to submit your entry for the AI Writing Contest, the submission deadline has been extended to December 16, 2024 . You now have two extra weeks to journey through the world of data collection for AI and LLM training for a shot at $2,500 in prizes. AI Writing Contest , AI Writing Contest submission deadline has been extended to December 16, 2024 $2,500 in prizes. Enter Now! Choose a writing template from the list below, fill it out, and submit it. The Future of Data Collection for AI and LLMs
Using Bright Data for AI and LLM Training
Web Scraping for AI and LLM Training Data Enter Now! Enter Now! Choose a writing template from the list below, fill it out, and submit it. The Future of Data Collection for AI and LLMs Using Bright Data for AI and LLM Training Web Scraping for AI and LLM Training Data The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training Web Scraping for AI and LLM Training Data Web Scraping for AI and LLM Training Data Web Scraping for AI and LLM Training Data AI is everywhere—from kids using it at school to professionals relying on it in their work, and even in your mom's Facebook group chat generating stickers. This contest shifts the focus from the flashy front end most users are familiar with to the unsung hero: data. data. Since we launched in September, participants have shared unique insights on data collection for AI and LLM training. For example, @ayinketh discusses overcoming traditional web scraping challenges with Bright Data in his entry , while @adnanaleeza uncovers a sentiment analysis hack in hers. Others explored everything from AI hallucinations to industry analysis. @ayinketh his entry his entry @adnanaleeza a sentiment analysis hack a sentiment analysis hack With just two weeks left and $2,500 still up for grabs , now’s the perfect time to enter the contest. With just two weeks left and $2,500 still up for grabs Tips for Entering the AI Writing Contest 1. Learn About Bright Data 1. Learn About Bright Data Start by watching the Introduction to Bright Datasets video below to understand how Bright Data works. Introduction to Bright Datasets Bright Data https://youtu.be/sUDROEVLqEk?feature=shared&embedable=true https://youtu.be/sUDROEVLqEk?feature=shared&embedable=true 2. Explore Bright Data 2. Explore Bright Data Take advantage of Bright Data’s $15 credit to explore its Dataset Marketplace and Web Data for AI. It’s a great way to get hands-on experience with the tools you’ll be using. Take advantage of Bright Data’s $15 credit Dataset Marketplace Dataset Marketplace Web Data for AI . Web Data for AI 3. Fill Out a Contest Template 3. Fill Out a Contest Template Choose a template from the list below, fill it out, and submit your entry. The Future of Data Collection for AI and LLMs Using Bright Data for AI and LLM Training Web Scraping for AI and LLM Training Data The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training Web Scraping for AI and LLM Training Data Web Scraping for AI and LLM Training Data Web Scraping for AI and LLM Training Data That’s all! By following these steps, you could win any of the prizes below: By following these steps, you could win any of the prizes below: Prize Breakdown for the AI Writing Contest

$$$



Best Bright Data Story

$1000



Best AI Story

$1000



Best Web Scraping Story

$500 Prize Breakdown for the AI Writing Contest

$$$



Best Bright Data Story

$1000



Best AI Story

$1000



Best Web Scraping Story

$500 Prize Breakdown for the AI Writing Contest

$$$ Best Bright Data Story

$1000 Best AI Story

$1000 Best Web Scraping Story

$500 Prize Breakdown for the AI Writing Contest $$$ Prize Breakdown for the AI Writing Contest Prize Breakdown for the AI Writing Contest Prize Breakdown for the AI Writing Contest $$$ $$$ $$$ Best Bright Data Story $1000 Best Bright Data Story Best Bright Data Story $1000 $1000 Best AI Story $1000 Best AI Story Best AI Story $1000 $1000 Best Web Scraping Story $500 Best Web Scraping Story Best Web Scraping Story $500 $500 Rules & Guidelines Must be 18+ to enter (no location restrictions) Create a HackerNoon account You can enter the contest by adding the #ai tag to your story. No AI-generated entries. (Read: the acceptable way to use AI on HackerNoon) Must be 18+ to enter (no location restrictions) Create a HackerNoon account Create a HackerNoon account You can enter the contest by adding the #ai tag to your story. #ai No AI-generated entries. (Read: the acceptable way to use AI on HackerNoon ) the acceptable way to use AI on HackerNoon FAQs Can I Write Under a Pen Name? Yes! You can use your real name, or a pseudonym when setting up your HackerNoon profile. How Long Will the Contest Run? The contest is open until December 16, 2024. December 16, 2024. How are the winners selected? After the contest ends, we’ll review submitted entries and shortlist the stories that receive the most eyeballs (real humans, not bots!). Next, the shortlisted stories will be voted on by HackerNoon staff. Then, the best in each prize category; AI, Bright Data, and Web Scraping, will be selected and announced After the contest ends, we’ll review submitted entries and shortlist the stories that receive the most eyeballs (real humans, not bots!). Next, the shortlisted stories will be voted on by HackerNoon staff. Then, the best in each prize category; AI, Bright Data, and Web Scraping, will be selected and announced Ready to enter the AI writing contest? Ready to enter the AI writing contest? Start a draft now, or enter the contest using any of these templates: Using Bright Data for AI and LLM Training
The Future of Data Collection for AI and LLMs
Web Scraping for AI and LLM Data Collection Start a draft now , or enter the contest using any of these templates: Start a draft now Start a draft now Start a draft now Using Bright Data for AI and LLM Training The Future of Data Collection for AI and LLMs Web Scraping for AI and LLM Data Collection Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training Using Bright Data for AI and LLM Training The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs The Future of Data Collection for AI and LLMs Web Scraping for AI and LLM Data Collection Web Scraping for AI and LLM Data Collection Web Scraping for AI and LLM Data Collection Web Scraping for AI and LLM Data Collection Good luck! Good luck!

Deadline Extended: 2 Weeks Left to Compete for $2,500 in the AI Writing Contest 

Table of Links Abstract and Introduction
Related Work
Problem Definition
Method
Experiments
Conclusion and References A. Appendix A.1. Full Prompts and A.2 ICPL Details A. 3 Baseline Details A.4 Environment Details A.5 Proxy Human Preference A.6 Human-in-the-Loop Preference A.4 ENVIRONMENT DETAILS In Table 4, we present the observation and action dimensions, along with the task description and task metrics for 9 tasks in IsaacGym. Authors:
(1) Chao Yu, Tsinghua University;
(2) Hong Lu, Tsinghua University;
(3) Jiaxuan Gao, Tsinghua University;
(4) Qixin Tan, Tsinghua University;
(5) Xinting Yang, Tsinghua University;
(6) Yu Wang, with equal advising from Tsinghua University;
(7) Yi Wu, with equal advising from Tsinghua University and the Shanghai Qi Zhi Institute;
(8) Eugene Vinitsky, with equal advising from New York University (zoeyuchao@gmail.com). This paper is available on arxiv under CC 4.0 license. Table of Links Abstract and Introduction Related Work Problem Definition Method Experiments Conclusion and References Abstract and Introduction Abstract and Introduction Related Work Related Work Problem Definition Problem Definition Method Method Experiments Experiments Conclusion and References Conclusion and References A. Appendix A. Appendix A.1. Full Prompts and A.2 ICPL Details A.1. Full Prompts and A.2 ICPL Details A. 3 Baseline Details A. 3 Baseline Details A.4 Environment Details A.4 Environment Details A.5 Proxy Human Preference A.5 Proxy Human Preference A.6 Human-in-the-Loop Preference A.6 Human-in-the-Loop Preference A.4 ENVIRONMENT DETAILS In Table 4, we present the observation and action dimensions, along with the task description and task metrics for 9 tasks in IsaacGym. Authors: (1) Chao Yu, Tsinghua University; (2) Hong Lu, Tsinghua University; (3) Jiaxuan Gao, Tsinghua University; (4) Qixin Tan, Tsinghua University; (5) Xinting Yang, Tsinghua University; (6) Yu Wang, with equal advising from Tsinghua University; (7) Yi Wu, with equal advising from Tsinghua University and the Shanghai Qi Zhi Institute; (8) Eugene Vinitsky, with equal advising from New York University (zoeyuchao@gmail.com). Authors: Authors: (1) Chao Yu, Tsinghua University; (2) Hong Lu, Tsinghua University; (3) Jiaxuan Gao, Tsinghua University; (4) Qixin Tan, Tsinghua University; (5) Xinting Yang, Tsinghua University; (6) Yu Wang, with equal advising from Tsinghua University; (7) Yi Wu, with equal advising from Tsinghua University and the Shanghai Qi Zhi Institute; (8) Eugene Vinitsky, with equal advising from New York University ( zoeyuchao@gmail.com ). zoeyuchao@gmail.com This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Few-shot In-Context Preference Learning Using Large Language Models: Environment Details

Table of Links Abstract and Introduction
Related Work
Problem Definition
Method
Experiments
Conclusion and References A. Appendix A.1. Full Prompts and A.2 ICPL Details A. 3 Baseline Details A.4 Environment Details A.5 Proxy Human Preference A.6 Human-in-the-Loop Preference A APPENDIX We would suggest visiting https://sites.google.com/view/few-shot-icpl/home for more information and videos. A.1 FULL PROMPTS A.2 ICPL DETAILS The full pseudocode of ICPL is listed in Algo. 2. Authors:
(1) Chao Yu, Tsinghua University;
(2) Hong Lu, Tsinghua University;
(3) Jiaxuan Gao, Tsinghua University;
(4) Qixin Tan, Tsinghua University;
(5) Xinting Yang, Tsinghua University;
(6) Yu Wang, with equal advising from Tsinghua University;
(7) Yi Wu, with equal advising from Tsinghua University and the Shanghai Qi Zhi Institute;
(8) Eugene Vinitsky, with equal advising from New York University (zoeyuchao@gmail.com). This paper is available on arxiv under CC 4.0 license. Table of Links Abstract and Introduction Related Work Problem Definition Method Experiments Conclusion and References Abstract and Introduction Abstract and Introduction Related Work Related Work Problem Definition Problem Definition Method Method Experiments Experiments Conclusion and References Conclusion and References A. Appendix A. Appendix A.1. Full Prompts and A.2 ICPL Details A.1. Full Prompts and A.2 ICPL Details A. 3 Baseline Details A. 3 Baseline Details A.4 Environment Details A.4 Environment Details A.5 Proxy Human Preference A.5 Proxy Human Preference A.6 Human-in-the-Loop Preference A.6 Human-in-the-Loop Preference A APPENDIX We would suggest visiting https://sites.google.com/view/few-shot-icpl/home for more information and videos. A.1 FULL PROMPTS A.2 ICPL DETAILS The full pseudocode of ICPL is listed in Algo. 2. Authors: (1) Chao Yu, Tsinghua University; (2) Hong Lu, Tsinghua University; (3) Jiaxuan Gao, Tsinghua University; (4) Qixin Tan, Tsinghua University; (5) Xinting Yang, Tsinghua University; (6) Yu Wang, with equal advising from Tsinghua University; (7) Yi Wu, with equal advising from Tsinghua University and the Shanghai Qi Zhi Institute; (8) Eugene Vinitsky, with equal advising from New York University (zoeyuchao@gmail.com). Authors: Authors: (1) Chao Yu, Tsinghua University; (2) Hong Lu, Tsinghua University; (3) Jiaxuan Gao, Tsinghua University; (4) Qixin Tan, Tsinghua University; (5) Xinting Yang, Tsinghua University; (6) Yu Wang, with equal advising from Tsinghua University; (7) Yi Wu, with equal advising from Tsinghua University and the Shanghai Qi Zhi Institute; (8) Eugene Vinitsky, with equal advising from New York University ( zoeyuchao@gmail.com ). zoeyuchao@gmail.com This paper is available on arxiv under CC 4.0 license. This paper is available on arxiv under CC 4.0 license. available on arxiv

Few-shot In-Context Preference Learning Using Large Language Models: Full Prompts and ICPL Details

#AI-chatbot Writing Contest Sponsor, Coze, Shares Secrets for Success as Contest Deadline Approaches

Coze, спонсар пісьменніцкага конкурсу #AI-chatbot, дзеліцца сакрэтамі поспеху па меры набліжэння дэдлайну конкурсу

#AI-chatbot Sponzor takmičenja u pisanju, Coze, dijeli tajne uspjeha kako se približava rok za takmičenje

El patrocinador del concurso de redacción de chatbots con inteligencia artificial, Coze, comparte sus secretos para el éxito a medida que se acerca la fecha límite del concurso

ربات #AI-chat Writing حامی مسابقه، Coze، اسرار موفقیت را با نزدیک شدن به آخرین مهلت مسابقه به اشتراک می گذارد

#AI-chatbot patwone konkou ekriti, Coze, pataje sekrè pou siksè pandan dat limit konkou a ap apwoche

#AI-chatbot íróverseny szponzora, Coze, megosztja a siker titkait a verseny határidejének közeledtével

#AI チャットボット ライティング コンテストのスポンサー Coze が、コンテストの締め切りが近づく中、成功の秘訣を披露

#AI-챗봇 글쓰기 대회 스폰서, 코즈, 대회 마감일이 다가오면서 성공 비결 공유

#AI-chatbot Sponsor al concursului de scriere, Coze, împărtășește secretele succesului pe măsură ce se apropie termenul limită al concursului

Спонсор письменницького конкурсу #AI-chatbot Coze ділиться секретами успіху з наближенням дедлайну конкурсу

Nhà tài trợ cuộc thi viết #AI-chatbot, Coze, chia sẻ bí quyết thành công khi thời hạn cuộc thi đang đến gần

#AI 聊天机器人写作比赛赞助商 Coze 在比赛截止日期临近之际分享成功秘诀

What IS this Chatbot Anyway? My AI chatbot: Top Content Bot, is a companion for anyone who is just getting started or wants to break into the world of content creation and influencing. This chatbot simplifies the process of developing social media strategies, content ideas, and tips to better capture one’s desired niche and shorten the learning curve for new creators. Wanna See A Quick Demo? You can experience the Top Content Bot right here. It’s easy to use and a resounding success when I ran tests to see if the bot was versatile enough to give niche advice while also being able to gently start beginners off on their content creation journeys. A video recording of the bot in use can be found here. Behind the Scenes: The Top Content Bot The creation of the bot was quite easy even though I don’t know how to code! Using Coze’s straightforward bot creation dashboard, I was able to create the Top Content Bot, integrating high-end APIs, workflows, and databases at the click of a button. There was also a YouTube tutorial video on Coze’s channel that made the already seamless process even easier with detailed explanations about each option offered and its effects on the bot’s eventual outcome. Initially, I used a prompt saying, “You are a content strategy and idea generation bot. Your job is to help me make great content by generating ideas for my videos, blogs, and any other shared media”. Thanks to the tutorial video, I used the “Optimize” option which helped me detail exactly what I wanted the bot to do with written code, templates, and formats. From there, I added what databases my bot would draw knowledge from. I went with YouTube, Google Search, TikTok, and Chat GPT4. These databases are vast with the knowledge of user-influenced search engine prompts and video formats for the chatbot to draw inspiration from when serving users. I also added a voice assistant to read out its questions or answers in order to accommodate users with visibility impairments. Start To Finish To begin, I went to the ‘+’ sign on the top left of the Coze homepage. I entered the name of my bot and a short description, and I was taken to this page which allowed me to type in and optimize my prompt for the creation of the bot. I thought I was going to get confused for sure, but the Coze tutorial video was very helpful in clearing up any ambiguities. A Preview and debug section ensured that every step of the way, I could correct any discrepancies before publishing the AI agent. I’m very proud of how efficient this bot is after only one iteration. The chat box is responsive both in English and French, so it is available to more users looking for a way to easily tap into their creativity. I hope that this chatbot makes at least one content creator’s work a bit easier to stay on top of. In the future, I hope that this chat will be instrumental in developing content creators and helping them get into the game faster instead of wasting precious time doubting themselves or doing research and strategy that can be completed in minutes. I also hope future iterations of this bot will include storyboard generation and script writing optimization to enhance the quality of the content being created by target users. Conclusively, this was the easiest foray into no-code tech I have experienced. It has inspired me to try my hands at even more AI bot creation in the future. I believe that as the barrier to non-code writers in the internet community continues to be lowered, the quality of tech solutions available will begin to rise. What IS this Chatbot Anyway? My AI chatbot: Top Content Bot, is a companion for anyone who is just getting started or wants to break into the world of content creation and influencing. This chatbot simplifies the process of developing social media strategies, content ideas, and tips to better capture one’s desired niche and shorten the learning curve for new creators. Wanna See A Quick Demo? You can experience the Top Content Bot right here . It’s easy to use and a resounding success when I ran tests to see if the bot was versatile enough to give niche advice while also being able to gently start beginners off on their content creation journeys. here A video recording of the bot in use can be found here. A video recording of the bot in use can be found here . here Behind the Scenes: The Top Content Bot The creation of the bot was quite easy even though I don’t know how to code! Using Coze’s straightforward bot creation dashboard, I was able to create the Top Content Bot, integrating high-end APIs, workflows, and databases at the click of a button. There was also a YouTube tutorial video on Coze’s channel that made the already seamless process even easier with detailed explanations about each option offered and its effects on the bot’s eventual outcome. Initially, I used a prompt saying, “You are a content strategy and idea generation bot. Your job is to help me make great content by generating ideas for my videos, blogs, and any other shared media”. Thanks to the tutorial video, I used the “Optimize” option which helped me detail exactly what I wanted the bot to do with written code, templates, and formats. From there, I added what databases my bot would draw knowledge from. I went with YouTube, Google Search, TikTok, and Chat GPT4. These databases are vast with the knowledge of user-influenced search engine prompts and video formats for the chatbot to draw inspiration from when serving users. I also added a voice assistant to read out its questions or answers in order to accommodate users with visibility impairments. Start To Finish To begin, I went to the ‘+’ sign on the top left of the Coze homepage. I entered the name of my bot and a short description, and I was taken to this page which allowed me to type in and optimize my prompt for the creation of the bot. I thought I was going to get confused for sure, but the Coze tutorial video was very helpful in clearing up any ambiguities. A Preview and debug section ensured that every step of the way, I could correct any discrepancies before publishing the AI agent. I’m very proud of how efficient this bot is after only one iteration. The chat box is responsive both in English and French, so it is available to more users looking for a way to easily tap into their creativity. I hope that this chatbot makes at least one content creator’s work a bit easier to stay on top of. In the future, I hope that this chat will be instrumental in developing content creators and helping them get into the game faster instead of wasting precious time doubting themselves or doing research and strategy that can be completed in minutes. I also hope future iterations of this bot will include storyboard generation and script writing optimization to enhance the quality of the content being created by target users. Conclusively, this was the easiest foray into no-code tech I have experienced. It has inspired me to try my hands at even more AI bot creation in the future. I believe that as the barrier to non-code writers in the internet community continues to be lowered, the quality of tech solutions available will begin to rise.

Struggling with Content Creation? I Built an AI Chatbot to Help You 

I was wrong. 18 months ago, I launched @storyarb to be the best content agency for founders & executives. My vision was a $10,000,000/year ghostwriting business. SPOILER: we pivoted. But the future has never looked brighter. Today, storyarb is a B2B company's outsourced Head of Content Marketing. We help our partners create high-quality, niche content for their ICP, while driving scalable demand & pipeline. Keep reading (below) for the full story... I launched storyarb to be the best content agency for founders & executives. I saw the acceleration of short-form video agencies (viralcuts, clipt, shortzy), but felt like there were no businesses supporting execs on text-based social (twitter & linkedin). From my own experience (and that of other founder friends), I knew that Twitter/Linkedin could be a magnet of opportunity. From my own experience (and that of other founder friends), I knew that Twitter/Linkedin could be a magnet of opportunity. It’s how I’ve met amazing entrepreneurs like @RomeenSheth , @jspujji , @alexisohanian.
It’s how I’ve gotten 8-figures worth of advertisers at @MorningBrew.
And it’s how I’ve worked through business ideas, company challenges, & thoughts on where the world is going. At the same time, I knew most founders didn’t have the chops or the time to create great content, solo on social. Which led to a literal shower thought. What if I pair founders with world-class strategists & copywriters. A strategist would pull unique insights from the founder & help set their social strategy. A copywriter would translate ideas into sharp content. Seems logical, right? Personal brand is a nice-to-have, not a need-to-have for founders. The result was 5-6 months of smooth sailing with a client followed by the gutpunch… “Can we catch up?” Churn for exec social became unpalatable. It was driven by: Founders becoming impatient & expecting results sooner.
Founders being pulled in 50 directions & lacking time for social content.
Founders feeling like their internet voice was slightly different from their real life voice. My cofounder, @AbbyMurray, and I knew that we had to get off the hamster wheel if we wanted to build a $10m biz. So we did two things: First, we asked ourselves: “What unfair advantages do we have as a business today?” “The best writers & strategists on the internet, period. We have better ideas & execution.” Second, we listened to our customers (and honestly, no one else). We reflected on past convos: A B2B HR Tech co asked if we could write their newsletter
An e-comm SaaS founder asked for help with e-books
A CMO wanted a company social revival We also interviewed the shit out of our customers. We asked them if they had any other content needs. We asked their marketing teams about their biggest current challenges. We asked “if you had access to world-class writers/content minds, how would you use them?” The answer (and our pivot) became clear. Founders kinda need content help. B2B companies reallyyyy need content help. Marketing teams are scrambling to stand out & grow their business today. Marketing costs are rising (adios cookies, bye zero interest rates, hello increased competition)
Product differentiation is narrowing (software + AI begets more software + AI)
Attribution has become religion, which forces marketers into a box of undifferentiated content
Most brands are living in a sea of sameness (groupthink + AI tooling)
Headcount is recalibrating (we no longer live in pre-COVID La La Land) Which means every marketer & every company is looking to stand out creatively & acquire customers cheaply & scalably. High-quality, high-value content that drives demand is one of the answers. But marketing orgs aren’t equipped to build & execute stand-out content strategies for 5 main reasons: They don’t know how to hire great content talent (copywriters, video creators, producers, strategists), because they’ve never had to do so.
They have marketers doing a lot of the content work, but are starting to realize it’s a completely different skill from most marketing responsibilities.
They have 1-2 team members creating content, but it feels flat & they have no idea what the hell to do now.
They do a poor job of tracking performance & connecting content to pipeline.
Content isn’t centralized...so it’s good in some places, fine in others, and f*cking atrocious elsewhere. Today, storyarb is your outsourced Head of Content Marketing, helping to resuscitate your company’s lifeless content (no matter which situation you’re in). It starts with us understanding the state of your content today. This includes a literal content colonoscopy up & down the org: Newsletters
Email marketing
Company social
Founder social
Ad creative
Website copy
Blog posts
Lead magnets
SOPs/technical docs
[FILL IN THE BLANK] Next, our strategists & writers work in lockstep to UNF*CK YOUR CONTENT by answering one crucial question: “What strategy (read: channels & ideas) will allow us to drive scalable customer demand without sacrificing the quality of our content?” That was the pivot. And it’s working in a big way. Series C tech startups to publicly traded multinationals trust us to be their Head of Content Marketing & we’re driving pipeline with high-quality, niche content built for our customers’ ICP. A few final notes as I reflect on the experience & journey building @storyarb alongside Abby over the last year. Pivots are basically inevitable. The odds that your day 1 product is your day 365 product is close to zero. It’s impossible to make perfect assumptions pre-launch & what’s most important is how you sift through info from your customers & make intuitive & informed decisions thereafter.


There is a business to build around founder social, it’s just not a recurring revenue agency. If I was to build a founder/exec social-only business again, it’d be a combo of courses (organized by platform & expertise), corporate workshops & paid community (accountability/engagement groups).


Building a great business is far more important than being right. As a first-time founder, you want to stay the course & protect your ego. As a second-time founder, you want to make adjustments & build a strong, dependable business.


Great talent creates great competencies. Great competencies create options. The only reason we’ve been able to evolve storyarb & crush it for our current clients is because the business was built on a foundation of exceptional talent. I hope this post made pivoting feel a little less like a failure to you. And if you’re a marketer, I hope this post made you feel like you don’t have to settle for mediocre content, while growing your business. I was wrong. 18 months ago, I launched @storyarb to be the best content agency for founders & executives. My vision was a $10,000,000/year ghostwriting business. @storyarb SPOILER: we pivoted. But the future has never looked brighter. Today, storyarb is a B2B company's outsourced Head of Content Marketing. We help our partners create high-quality, niche content for their ICP, while driving scalable demand & pipeline. Keep reading (below) for the full story... I launched storyarb to be the best content agency for founders & executives. I saw the acceleration of short-form video agencies (viralcuts, clipt, shortzy), but felt like there were no businesses supporting execs on text-based social (twitter & linkedin). From my own experience (and that of other founder friends), I knew that Twitter/Linkedin could be a magnet of opportunity. From my own experience (and that of other founder friends), I knew that Twitter/Linkedin could be a magnet of opportunity. It’s how I’ve met amazing entrepreneurs like @RomeenSheth , @jspujji , @alexisohanian. It’s how I’ve gotten 8-figures worth of advertisers at @MorningBrew. And it’s how I’ve worked through business ideas, company challenges, & thoughts on where the world is going. It’s how I’ve met amazing entrepreneurs like @RomeenSheth , @jspujji , @alexisohanian . @RomeenSheth @jspujji @alexisohanian It’s how I’ve gotten 8-figures worth of advertisers at @MorningBrew . @MorningBrew And it’s how I’ve worked through business ideas, company challenges, & thoughts on where the world is going. At the same time, I knew most founders didn’t have the chops or the time to create great content, solo on social. Which led to a literal shower thought. What if I pair founders with world-class strategists & copywriters. A strategist would pull unique insights from the founder & help set their social strategy. A copywriter would translate ideas into sharp content. Seems logical, right? Personal brand is a nice-to-have, not a need-to-have for founders. The result was 5-6 months of smooth sailing with a client followed by the gutpunch… “Can we catch up?” “Can we catch up?” Churn for exec social became unpalatable. It was driven by: Founders becoming impatient & expecting results sooner. Founders being pulled in 50 directions & lacking time for social content. Founders feeling like their internet voice was slightly different from their real life voice. Founders becoming impatient & expecting results sooner. Founders being pulled in 50 directions & lacking time for social content. Founders feeling like their internet voice was slightly different from their real life voice. My cofounder, @AbbyMurray , and I knew that we had to get off the hamster wheel if we wanted to build a $10m biz. So we did two things: @AbbyMurray First, we asked ourselves: “What unfair advantages do we have as a business today?” “The best writers & strategists on the internet, period. We have better ideas & execution.” Second, we listened to our customers (and honestly, no one else). We reflected on past convos: A B2B HR Tech co asked if we could write their newsletter An e-comm SaaS founder asked for help with e-books A CMO wanted a company social revival A B2B HR Tech co asked if we could write their newsletter An e-comm SaaS founder asked for help with e-books A CMO wanted a company social revival We also interviewed the shit out of our customers. We asked them if they had any other content needs. We asked their marketing teams about their biggest current challenges. We asked “if you had access to world-class writers/content minds, how would you use them?” The answer (and our pivot) became clear. Founders kinda need content help. B2B companies reallyyyy need content help. Marketing teams are scrambling to stand out & grow their business today. Marketing costs are rising (adios cookies, bye zero interest rates, hello increased competition) Product differentiation is narrowing (software + AI begets more software + AI) Attribution has become religion, which forces marketers into a box of undifferentiated content Most brands are living in a sea of sameness (groupthink + AI tooling) Headcount is recalibrating (we no longer live in pre-COVID La La Land) Marketing costs are rising (adios cookies, bye zero interest rates, hello increased competition) Product differentiation is narrowing (software + AI begets more software + AI) Attribution has become religion, which forces marketers into a box of undifferentiated content Most brands are living in a sea of sameness (groupthink + AI tooling) Headcount is recalibrating (we no longer live in pre-COVID La La Land) Which means every marketer & every company is looking to stand out creatively & acquire customers cheaply & scalably. High-quality, high-value content that drives demand is one of the answers. But marketing orgs aren’t equipped to build & execute stand-out content strategies for 5 main reasons: They don’t know how to hire great content talent (copywriters, video creators, producers, strategists), because they’ve never had to do so. They have marketers doing a lot of the content work, but are starting to realize it’s a completely different skill from most marketing responsibilities. They have 1-2 team members creating content, but it feels flat & they have no idea what the hell to do now. They do a poor job of tracking performance & connecting content to pipeline. Content isn’t centralized...so it’s good in some places, fine in others, and f*cking atrocious elsewhere. They don’t know how to hire great content talent (copywriters, video creators, producers, strategists), because they’ve never had to do so. They have marketers doing a lot of the content work, but are starting to realize it’s a completely different skill from most marketing responsibilities. They have 1-2 team members creating content, but it feels flat & they have no idea what the hell to do now. They do a poor job of tracking performance & connecting content to pipeline. Content isn’t centralized...so it’s good in some places, fine in others, and f*cking atrocious elsewhere. Today, storyarb is your outsourced Head of Content Marketing, helping to resuscitate your company’s lifeless content (no matter which situation you’re in). It starts with us understanding the state of your content today. This includes a literal content colonoscopy up & down the org: Newsletters Email marketing Company social Founder social Ad creative Website copy Blog posts Lead magnets SOPs/technical docs [FILL IN THE BLANK] Newsletters Email marketing Company social Founder social Ad creative Website copy Blog posts Lead magnets SOPs/technical docs [FILL IN THE BLANK] Next, our strategists & writers work in lockstep to UNF*CK YOUR CONTENT by answering one crucial question: “What strategy (read: channels & ideas) will allow us to drive scalable customer demand without sacrificing the quality of our content?” “What strategy (read: channels & ideas) will allow us to drive scalable customer demand without sacrificing the quality of our content?” That was the pivot. And it’s working in a big way. Series C tech startups to publicly traded multinationals trust us to be their Head of Content Marketing & we’re driving pipeline with high-quality, niche content built for our customers’ ICP. A few final notes as I reflect on the experience & journey building @storyarb alongside Abby over the last year. @storyarb Pivots are basically inevitable. The odds that your day 1 product is your day 365 product is close to zero. It’s impossible to make perfect assumptions pre-launch & what’s most important is how you sift through info from your customers & make intuitive & informed decisions thereafter. There is a business to build around founder social, it’s just not a recurring revenue agency. If I was to build a founder/exec social-only business again, it’d be a combo of courses (organized by platform & expertise), corporate workshops & paid community (accountability/engagement groups). Building a great business is far more important than being right. As a first-time founder, you want to stay the course & protect your ego. As a second-time founder, you want to make adjustments & build a strong, dependable business. Great talent creates great competencies. Great competencies create options. The only reason we’ve been able to evolve storyarb & crush it for our current clients is because the business was built on a foundation of exceptional talent. Pivots are basically inevitable. The odds that your day 1 product is your day 365 product is close to zero. It’s impossible to make perfect assumptions pre-launch & what’s most important is how you sift through info from your customers & make intuitive & informed decisions thereafter. Pivots are basically inevitable. The odds that your day 1 product is your day 365 product is close to zero. It’s impossible to make perfect assumptions pre-launch & what’s most important is how you sift through info from your customers & make intuitive & informed decisions thereafter. There is a business to build around founder social, it’s just not a recurring revenue agency. If I was to build a founder/exec social-only business again, it’d be a combo of courses (organized by platform & expertise), corporate workshops & paid community (accountability/engagement groups). There is a business to build around founder social, it’s just not a recurring revenue agency. If I was to build a founder/exec social-only business again, it’d be a combo of courses (organized by platform & expertise), corporate workshops & paid community (accountability/engagement groups). Building a great business is far more important than being right. As a first-time founder, you want to stay the course & protect your ego. As a second-time founder, you want to make adjustments & build a strong, dependable business. Building a great business is far more important than being right. As a first-time founder, you want to stay the course & protect your ego. As a second-time founder, you want to make adjustments & build a strong, dependable business. Great talent creates great competencies. Great competencies create options. The only reason we’ve been able to evolve storyarb & crush it for our current clients is because the business was built on a foundation of exceptional talent. Great talent creates great competencies. Great competencies create options. The only reason we’ve been able to evolve storyarb & crush it for our current clients is because the business was built on a foundation of exceptional talent. I hope this post made pivoting feel a little less like a failure to you. And if you’re a marketer, I hope this post made you feel like you don’t have to settle for mediocre content, while growing your business.

Why storyarb Pivoted From Ghostwriting for Founders to Producing Niche Content for Customers’ ICP

Proč storyarb přešel od Ghostwritingu pro zakladatele k vytváření specializovaného obsahu pro ICP zákazníků

Por qué Storyarb pasó de ser redactora fantasma para fundadores a producir contenido de nicho para el ICP de los clientes

Miksi storyarb siirtyi haamukirjoituksesta perustajille niche-sisällön tuottamiseen asiakkaiden ICP:lle

Mengapa storyarb Beralih Dari Ghostwriting untuk Pendiri ke Produksi Konten Niche untuk Pelanggan

storyarb が創業者向けのゴーストライター業務から顧客の ICP 向けニッチ コンテンツの制作に方向転換した理由

Storyarb가 창업자를 위한 고스트라이터에서 고객의 ICP를 위한 틈새 콘텐츠 제작으로 전환한 이유

Mpo na nini storyarb Pivoted De Ghostwriting mpo na ba Fondateurs mpo na kobimisa Niche Content mpo na Clients’ ICP

Nahoana ny storyarb no nivadika avy amin'ny Ghostwriting ho an'ny mpanorina mba hamokarana votoaty manokana ho an'ny ICP mpanjifa

Зошто storyarb се сврти од Ghostwriting за основачи до производство на ниша содржина за ICP на клиентите

किन स्टोरीआर्ब संस्थापकहरूको लागि भूतलेखनबाट ग्राहकहरूको आईसीपीको लागि आला सामग्री उत्पादन गर्न पिभोट गरियो

Чаро storyarb аз Ghostwriting барои муассисон ба истеҳсоли мундариҷаи нишеӣ барои ICP муштариён табдил ёфт

Чому storyarb перейшов від Ghostwriting для засновників до створення нішевого контенту для ICP клієнтів

Let’s talk about something that’s been driving me nuts for years: Google’s ongoing struggles with handling canonical URLs and duplicate content. It’s a problem that, despite being fundamental to search engine optimization (SEO), Google still can’t seem to get right. And while Google continues to flounder, guess who’s quietly acing it? Bing. Yes, you read that right – Bing! It’s honestly baffling. Here we are, in 2024, and Google, the self-proclaimed king of search, can’t even get the basics of web indexing straight. Let’s break this down. The Canonical URL Catastrophe A canonical URL is meant to tell search engines which version of a page is the “official” one when you have multiple URLs with similar or identical content. Seems simple enough, right? You’d think Google, with all its technological might and years of search engine dominance, would have perfected this by now. But nope. Google’s handling of canonical tags is, more often than not, a dumpster fire. Why? Because Google has a tendency to ignore your carefully placed canonical tags, or worse, it just gets them wrong. It might pick the wrong version of your page, or even worse – it might randomly declare some other URL as the canonical without any explanation. This leads to pages getting improperly indexed, with duplicate content spread across multiple URLs and absolutely no way for you to control it. Result? Your rankings plummet, your SEO strategy crumbles, and your site’s traffic tanks – all thanks to Google’s refusal to play by the canonical rulebook. But, oh, here’s where it gets good. Bing? Bing doesn’t have these problems. Bing can crawl your site, see your canonical tag, and follow it like a well-behaved child. It doesn’t second-guess you, doesn’t ignore your instructions, and doesn’t make arbitrary decisions that hurt your visibility. Bing understands that when a site owner points to a canonical URL, that’s what it should respect. It’s as if Bing actually listens to website owners, while Google seems to be playing SEO bingo with your site’s content. Duplicate Content Woes Then there’s duplicate content. Another area where Google’s “genius” just falls flat. We all know duplicate content is a big issue when it comes to SEO. When search engines find multiple copies of the same content, they’re supposed to figure out which one is the most relevant and display it in search results. Seems simple, right? Well, it’s not. Google has been notorious for misidentifying duplicate content, sometimes even penalizing sites with pages that have only slight variations. I’ve seen it happen time and again: you publish a well-researched article, and somehow, Google finds another page on your site and decides they’re too similar, tanking your rankings. Why does this happen? It’s because Google’s algorithm, despite being incredibly complex, still can’t reliably handle duplicate content in certain situations. There are cases where Google just can’t seem to distinguish between genuinely different pages and what’s essentially the same content presented slightly differently – even if those pages are clearly differentiated by their canonical tags or intended purpose. The result? Duplicate content penalties, frustrated webmasters, and a whole lot of wasted time trying to get Google to "understand" your site. Now, Bing? Bing doesn’t overcomplicate things. It doesn’t see minor content variations and freak out like Google does. It doesn’t penalize you for every slight difference between two pages that have the same general topic. Instead, Bing takes a more balanced approach to identifying and handling duplicate content. You won’t find yourself scrambling to fix duplicate content issues that Google invented in the first place. It just works. The Moral of the Story Why is Bing, a search engine that many still dismiss, consistently getting the basics right while Google is out here making life unnecessarily difficult for website owners? The answer is simple: Google is overcomplicating things, and it’s failing to acknowledge the importance of consistency and accuracy in indexing and content management. In contrast, Bing seems to have a more grounded, practical approach. Bing might not dominate the search engine market the way Google does, but at least when it comes to handling canonical URLs and duplicate content, Bing is getting it right. It’s time for Google to stop pretending that its “complex algorithms” are the answer to everything. Sometimes, simplicity works better. Maybe it’s time for Google to take a page out of Bing’s playbook and fix the very issues that have been plaguing SEO professionals for years. Because if Bing can get it right, there’s absolutely no excuse for Google to keep getting it wrong. Let’s talk about something that’s been driving me nuts for years: Google’s ongoing struggles with handling canonical URLs and duplicate content. It’s a problem that, despite being fundamental to search engine optimization (SEO), Google still can’t seem to get right. And while Google continues to flounder, guess who’s quietly acing it? Bing. Yes, you read that right – Bing! And while Google continues to flounder, guess who’s quietly acing it? Bing. Yes, you read that right – Bing! It’s honestly baffling. Here we are, in 2024, and Google, the self-proclaimed king of search, can’t even get the basics of web indexing straight. Let’s break this down. The Canonical URL Catastrophe A canonical URL is meant to tell search engines which version of a page is the “official” one when you have multiple URLs with similar or identical content. Seems simple enough, right? You’d think Google, with all its technological might and years of search engine dominance, would have perfected this by now. But nope. Google’s handling of canonical tags is, more often than not, a dumpster fire. Why? Because Google has a tendency to ignore your carefully placed canonical tags, or worse, it just gets them wrong. It might pick the wrong version of your page, or even worse – it might randomly declare some other URL as the canonical without any explanation. ignore randomly This leads to pages getting improperly indexed, with duplicate content spread across multiple URLs and absolutely no way for you to control it. Result? Your rankings plummet, your SEO strategy crumbles, and your site’s traffic tanks – all thanks to Google’s refusal to play by the canonical rulebook. SEO strategy But, oh, here’s where it gets good. Bing ? Bing doesn’t have these problems. Bing can crawl your site, see your canonical tag, and follow it like a well-behaved child. It doesn’t second-guess you, doesn’t ignore your instructions, and doesn’t make arbitrary decisions that hurt your visibility. Bing Bing understands that when a site owner points to a canonical URL, that’s what it should respect. It’s as if Bing actually listens to website owners, while Google seems to be playing SEO bingo with your site’s content. Duplicate Content Woes Then there’s duplicate content. Another area where Google’s “genius” just falls flat. We all know duplicate content is a big issue when it comes to SEO. When search engines find multiple copies of the same content, they’re supposed to figure out which one is the most relevant and display it in search results. Seems simple, right? Well, it’s not. Google has been notorious for misidentifying duplicate content, sometimes even penalizing sites with pages that have only slight variations. I’ve seen it happen time and again: you publish a well-researched article, and somehow, Google finds another page on your site and decides they’re too similar, tanking your rankings. Why does this happen? It’s because Google’s algorithm, despite being incredibly complex, still can’t reliably handle duplicate content in certain situations. There are cases where Google just can’t seem to distinguish between genuinely different pages and what’s essentially the same content presented slightly differently – even if those pages are clearly differentiated by their canonical tags or intended purpose. The result? Duplicate content penalties, frustrated webmasters, and a whole lot of wasted time trying to get Google to "understand" your site. Now, Bing? Bing doesn’t overcomplicate things. It doesn’t see minor content variations and freak out like Google does. It doesn’t penalize you for every slight difference between two pages that have the same general topic. Instead, Bing takes a more balanced approach to identifying and handling duplicate content. You won’t find yourself scrambling to fix duplicate content issues that Google invented in the first place. It just works. The Moral of the Story Why is Bing, a search engine that many still dismiss, consistently getting the basics right while Google is out here making life unnecessarily difficult for website owners? The answer is simple: Google is overcomplicating things, and it’s failing to acknowledge the importance of consistency and accuracy in indexing and content management . Google is overcomplicating things, and it’s failing to acknowledge the importance of consistency and accuracy in indexing and content management In contrast, Bing seems to have a more grounded, practical approach. Bing might not dominate the search engine market the way Google does, but at least when it comes to handling canonical URLs and duplicate content, Bing is getting it right. It’s time for Google to stop pretending that its “complex algorithms” are the answer to everything. Sometimes, simplicity works better. Maybe it’s time for Google to take a page out of Bing’s playbook and fix the very issues that have been plaguing SEO professionals for years. Because if Bing can get it right, there’s absolutely no excuse for Google to keep getting it wrong.

Stories, News & Mentions about CONTE

Conte’s Stories on HackerNoon

Conte's News & Mentions