Table of Contents
Network effects work differently for AI
OpenAI and Microsoft
Home Technology peripherals AI How network effects can make artificial intelligence smarter

How network effects can make artificial intelligence smarter

May 06, 2023 pm 12:10 PM
AI


How network effects can make artificial intelligence smarter

Network effects determine the success of technologies from phone calls to online shopping platforms, and AI tools like ChatGPT are no exception . The difference, however, is how these network effects work. Data network effects are a new form, like the more familiar direct and indirect network effects, where the value of a technology increases as the number of users it acquires increases.

However, the value we are discussing today is not the number of phone calls or the presence of many buyers and sellers on an e-commerce platform, but rather the feedback that comes from helping it make better predictions. More users means more responses, which further improves prediction accuracy, creating a virtuous cycle. There are three lessons for businesses to consider: 1) Feedback is critical, 2) Gather information regularly and meticulously, and 3) Consider the data that is shared intentionally or unintentionally.

When OpenAI launched ChatGPT late last year, the reaction from industry observers was a mixture of praise and concern. We’ve all heard how the technology is putting computer programmers, teachers, financial traders and analysts, graphic designers and artists alike out of work en masse. Many universities are rushing to revise course plans and related requirements out of concern that AI will stifle the creative abilities of college students. Others say that perhaps the most direct impact is that ChatGPT can reshape or even replace traditional Internet search engines. Search and related advertising generate the vast majority of Google’s revenue, so will chatbots be the end of Google?

ChatGPT is an excellent demonstration of machine learning technology, but is almost unfeasible as a standalone service. In order to leverage its technical strength, OpenAI needs a partner. So we weren't surprised when the company quickly announced a deal with Microsoft. The combination of an AI startup and an established tech company could ultimately pose a credible threat to Google's dominance, raising the stakes in the "AI arms race." It also provides a lesson on the forces that will determine which companies will thrive and which will falter when deploying this technology.

To understand what compelled OpenAI to align with Bing (and why Google might still win), we considered how the technology differs from past developments, such as phone calls or marketplace platforms like Uber or Airbnb. In these examples, network effects—where the value of a product rises as more users—play an important role in determining how these products grow and which companies succeed. Generative AI services like ChatGPT are subject to similar but different types of network effects. In order to choose an AI-friendly strategy, managers and entrepreneurs must understand how this new AI network effect works.

Network effects work differently for AI

The value of artificial intelligence lies in accurate predictions and recommendations. But unlike traditional products and services that rely on converting supplies (such as electricity or human capital) into outputs (such as lighting or tax advice), AI requires massive data sets that must be kept up to date through back-and-forth customer interactions. To remain competitive, AI operators must collect data, analyze the data, provide predictions, and then seek feedback to refine the recommendations. The value of the system depends on the data from the users and increases with the data.

The performance of this technology—the ability to accurately predict and advise—depends on an economic principle called data network effects (some prefer to call it data-driven learning). These are distinct from familiar direct network effects, such as a phone call becoming more valuable as your user base grows because you can call more people. They are also different from indirect network effects, which describe how more buyers invite more sellers to join the platform, and vice versa—when more sellers are present, shopping on an e-commerce platform or on Airbnb Booking a room becomes more attractive.

Data network effects are a new form: Like the more familiar effect, the more users, the more valuable the technology becomes. But here, value does not come from the number of peers, nor from the presence of many buyers and sellers. Rather, these impacts stem from the nature of the technology: AI improves through reinforcement learning, prediction, and feedback. As intelligence increases, systems can make better predictions, enhance their usefulness, attract new users and retain existing users. More users means more responses, which further improves prediction accuracy, creating a virtuous cycle.

Take Google Maps as an example, which uses AI to recommend the fastest route to your destination. This ability depends on predicting true traffic patterns in alternative paths, which is achieved by leveraging data from many users. Here, data users are also suppliers, and the more people use Google Maps, the more historical and concurrent data it accumulates. With a wealth of data, Google can compare countless predictions with actual results: Did you arrive at the time the app predicted? To refine the predictions, the app also needs your impressions: How good was the description? As objective facts and subjective comments accumulate, network effects come into play. These effects improve predictions and increase the value of the application to users and to Google.

Once we understand how network effects drive artificial intelligence, we can imagine the new strategies this technology will require.

OpenAI and Microsoft

Let’s start with the marriage between OpenAI and Microsoft. When we beta-tested ChatGPT, we were impressed by its creative, human-like responses, but also realized it had its bottlenecks: It relies on massive amounts of data last collected in 2021, so don't ask about recent events or even weather. Worse, it lacks a robust feedback loop mechanism.

However, through a link with Microsoft, OpenAI found a way to test its predictions. Bing users' questions - and how they rate the answers - are critical to updating and improving ChatGPT. The next step, we imagine, is for Microsoft to feed the massive cloud of user data it maintains into the algorithm. While ChatGPT can digest countless Excel sheets, PowerPoint presentations, Word documents, and LinkedIn resumes, it will do a better job at recreating those files to the delight or horror of those in the office.

There are at least three broad lessons here.

First of all, feedback is crucial. The value of AI increases with continued user response. To stay smart, the algorithm needs a data stream of current user selections and past recommendation ratings. Without feedback, even the best engineering algorithms won't stay smart for long. As OpenAI realizes, even the most complex models need to be linked to constantly flowing data sources. AI entrepreneurs should be well aware of this.

Second, executives shouldcollect informationregularly and meticulously to maximize the impact. They should go through typical financial and operating records. Useful data is everywhere, both inside and outside the enterprise. They may come from interactions with buyers, suppliers, and colleagues. For example, retailers can track what consumers view, what they put into their shopping carts, and what they ultimately pay for. Cumulatively, these tiny details can dramatically improve an AI system’s predictions. Even uncommon data, including data over which a business has no control, may be worth collecting. Weather data helps Google Maps predict traffic conditions. Tracking the keywords recruiters use to search for resumes can help LinkedIn provide candidates with tips for success.

Finally, Everyone should consider the data they share either intentionally or unintentionally. Facts and feedback are critical to building better predictions, but the value of your data may be captured by others, and executives should consider which AI can benefit from the data they share (or allow access to). Sometimes, they should limit sharing. For example, when Uber drivers use the app Waze to navigate, they help Google estimate the frequency and length of ride-hailing trips. Such data could be invaluable as Google considers operating self-driving taxis.

Additionally, when brands like Adidas sell products on Amazon, the retail giant can estimate demand for different brands (such as compared to Nike) and categories, as well as buyer sensitivity to price. . The results could feed into competitors or benefit Amazon's private-label products. To combat this, executives can avoid third-party platforms or intermediaries. They can negotiate data access, and they can work to maintain direct contact with their customers. Sometimes, the best solution may be to have data owners bind and share in a data exchange, as banks have done when establishing ways to share credit data.

When you consider AI network effects, we can better understand the future of the technology. You can also see how these effects, like other network effects, make the rich even richer. The dynamics behind AI mean that first movers may be richly rewarded, while followers, no matter how fast, may be left behind. It also means that when one has access to AI algorithms and data streams, the advantages accrue over time and cannot be easily surpassed. For executives, entrepreneurs, policymakers, and everyone else, the best and worst of artificial intelligence are yet to come.

The above is the detailed content of How network effects can make artificial intelligence smarter. For more information, please follow other related articles on the PHP Chinese website!

Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn

Hot AI Tools

Undresser.AI Undress

Undresser.AI Undress

AI-powered app for creating realistic nude photos

AI Clothes Remover

AI Clothes Remover

Online AI tool for removing clothes from photos.

Undress AI Tool

Undress AI Tool

Undress images for free

Clothoff.io

Clothoff.io

AI clothes remover

AI Hentai Generator

AI Hentai Generator

Generate AI Hentai for free.

Hot Article

R.E.P.O. Energy Crystals Explained and What They Do (Yellow Crystal)
2 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
Repo: How To Revive Teammates
4 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
Hello Kitty Island Adventure: How To Get Giant Seeds
3 weeks ago By 尊渡假赌尊渡假赌尊渡假赌

Hot Tools

Notepad++7.3.1

Notepad++7.3.1

Easy-to-use and free code editor

SublimeText3 Chinese version

SublimeText3 Chinese version

Chinese version, very easy to use

Zend Studio 13.0.1

Zend Studio 13.0.1

Powerful PHP integrated development environment

Dreamweaver CS6

Dreamweaver CS6

Visual web development tools

SublimeText3 Mac version

SublimeText3 Mac version

God-level code editing software (SublimeText3)

Bytedance Cutting launches SVIP super membership: 499 yuan for continuous annual subscription, providing a variety of AI functions Bytedance Cutting launches SVIP super membership: 499 yuan for continuous annual subscription, providing a variety of AI functions Jun 28, 2024 am 03:51 AM

This site reported on June 27 that Jianying is a video editing software developed by FaceMeng Technology, a subsidiary of ByteDance. It relies on the Douyin platform and basically produces short video content for users of the platform. It is compatible with iOS, Android, and Windows. , MacOS and other operating systems. Jianying officially announced the upgrade of its membership system and launched a new SVIP, which includes a variety of AI black technologies, such as intelligent translation, intelligent highlighting, intelligent packaging, digital human synthesis, etc. In terms of price, the monthly fee for clipping SVIP is 79 yuan, the annual fee is 599 yuan (note on this site: equivalent to 49.9 yuan per month), the continuous monthly subscription is 59 yuan per month, and the continuous annual subscription is 499 yuan per year (equivalent to 41.6 yuan per month) . In addition, the cut official also stated that in order to improve the user experience, those who have subscribed to the original VIP

Context-augmented AI coding assistant using Rag and Sem-Rag Context-augmented AI coding assistant using Rag and Sem-Rag Jun 10, 2024 am 11:08 AM

Improve developer productivity, efficiency, and accuracy by incorporating retrieval-enhanced generation and semantic memory into AI coding assistants. Translated from EnhancingAICodingAssistantswithContextUsingRAGandSEM-RAG, author JanakiramMSV. While basic AI programming assistants are naturally helpful, they often fail to provide the most relevant and correct code suggestions because they rely on a general understanding of the software language and the most common patterns of writing software. The code generated by these coding assistants is suitable for solving the problems they are responsible for solving, but often does not conform to the coding standards, conventions and styles of the individual teams. This often results in suggestions that need to be modified or refined in order for the code to be accepted into the application

Seven Cool GenAI & LLM Technical Interview Questions Seven Cool GenAI & LLM Technical Interview Questions Jun 07, 2024 am 10:06 AM

To learn more about AIGC, please visit: 51CTOAI.x Community https://www.51cto.com/aigc/Translator|Jingyan Reviewer|Chonglou is different from the traditional question bank that can be seen everywhere on the Internet. These questions It requires thinking outside the box. Large Language Models (LLMs) are increasingly important in the fields of data science, generative artificial intelligence (GenAI), and artificial intelligence. These complex algorithms enhance human skills and drive efficiency and innovation in many industries, becoming the key for companies to remain competitive. LLM has a wide range of applications. It can be used in fields such as natural language processing, text generation, speech recognition and recommendation systems. By learning from large amounts of data, LLM is able to generate text

Can fine-tuning really allow LLM to learn new things: introducing new knowledge may make the model produce more hallucinations Can fine-tuning really allow LLM to learn new things: introducing new knowledge may make the model produce more hallucinations Jun 11, 2024 pm 03:57 PM

Large Language Models (LLMs) are trained on huge text databases, where they acquire large amounts of real-world knowledge. This knowledge is embedded into their parameters and can then be used when needed. The knowledge of these models is "reified" at the end of training. At the end of pre-training, the model actually stops learning. Align or fine-tune the model to learn how to leverage this knowledge and respond more naturally to user questions. But sometimes model knowledge is not enough, and although the model can access external content through RAG, it is considered beneficial to adapt the model to new domains through fine-tuning. This fine-tuning is performed using input from human annotators or other LLM creations, where the model encounters additional real-world knowledge and integrates it

To provide a new scientific and complex question answering benchmark and evaluation system for large models, UNSW, Argonne, University of Chicago and other institutions jointly launched the SciQAG framework To provide a new scientific and complex question answering benchmark and evaluation system for large models, UNSW, Argonne, University of Chicago and other institutions jointly launched the SciQAG framework Jul 25, 2024 am 06:42 AM

Editor |ScienceAI Question Answering (QA) data set plays a vital role in promoting natural language processing (NLP) research. High-quality QA data sets can not only be used to fine-tune models, but also effectively evaluate the capabilities of large language models (LLM), especially the ability to understand and reason about scientific knowledge. Although there are currently many scientific QA data sets covering medicine, chemistry, biology and other fields, these data sets still have some shortcomings. First, the data form is relatively simple, most of which are multiple-choice questions. They are easy to evaluate, but limit the model's answer selection range and cannot fully test the model's ability to answer scientific questions. In contrast, open-ended Q&A

SOTA performance, Xiamen multi-modal protein-ligand affinity prediction AI method, combines molecular surface information for the first time SOTA performance, Xiamen multi-modal protein-ligand affinity prediction AI method, combines molecular surface information for the first time Jul 17, 2024 pm 06:37 PM

Editor | KX In the field of drug research and development, accurately and effectively predicting the binding affinity of proteins and ligands is crucial for drug screening and optimization. However, current studies do not take into account the important role of molecular surface information in protein-ligand interactions. Based on this, researchers from Xiamen University proposed a novel multi-modal feature extraction (MFE) framework, which for the first time combines information on protein surface, 3D structure and sequence, and uses a cross-attention mechanism to compare different modalities. feature alignment. Experimental results demonstrate that this method achieves state-of-the-art performance in predicting protein-ligand binding affinities. Furthermore, ablation studies demonstrate the effectiveness and necessity of protein surface information and multimodal feature alignment within this framework. Related research begins with "S

Five schools of machine learning you don't know about Five schools of machine learning you don't know about Jun 05, 2024 pm 08:51 PM

Machine learning is an important branch of artificial intelligence that gives computers the ability to learn from data and improve their capabilities without being explicitly programmed. Machine learning has a wide range of applications in various fields, from image recognition and natural language processing to recommendation systems and fraud detection, and it is changing the way we live. There are many different methods and theories in the field of machine learning, among which the five most influential methods are called the "Five Schools of Machine Learning". The five major schools are the symbolic school, the connectionist school, the evolutionary school, the Bayesian school and the analogy school. 1. Symbolism, also known as symbolism, emphasizes the use of symbols for logical reasoning and expression of knowledge. This school of thought believes that learning is a process of reverse deduction, through existing

SK Hynix will display new AI-related products on August 6: 12-layer HBM3E, 321-high NAND, etc. SK Hynix will display new AI-related products on August 6: 12-layer HBM3E, 321-high NAND, etc. Aug 01, 2024 pm 09:40 PM

According to news from this site on August 1, SK Hynix released a blog post today (August 1), announcing that it will attend the Global Semiconductor Memory Summit FMS2024 to be held in Santa Clara, California, USA from August 6 to 8, showcasing many new technologies. generation product. Introduction to the Future Memory and Storage Summit (FutureMemoryandStorage), formerly the Flash Memory Summit (FlashMemorySummit) mainly for NAND suppliers, in the context of increasing attention to artificial intelligence technology, this year was renamed the Future Memory and Storage Summit (FutureMemoryandStorage) to invite DRAM and storage vendors and many more players. New product SK hynix launched last year

See all articles