One Article Review
| Source |  AlienVault Lab Blog |
|---|---|
| Identifiant | 8473954 |
| Date de publication | 2024-04-01 10:00:00 (vue: 2024-04-01 10:07:11) |
| Titre | AI - le bon, le mauvais et effrayant AI - The Good, Bad, and Scary |
| Texte | AI and machine learning (ML) optimizes processes by making recommendations for optimizing productivity, reducing cycles, and maximizing efficiency. AI also optimizes human capital by performing mundane & repetitive tasks 24x7 without the need for rest and minimizing human errors. There are numerous benefits as to how AI can benefit society. As much as AI can propel human progress forward, it can be consequential to our own detriment without proper guidance. We need to understand the risks and challenges that comes with AI. Growing your knowledge in the new era of AI will help you and your organization evolve. AI can be a battlefield of good and evil. There’s the power to do good and the power to do evil. Here are some examples on the Good, Bad, and Scary of AI. Good Cybersecurity - Detect and respond to cyber-attacks with automation capabilities at machine speed and predict behavioral anomalies and defend against cyber threats before an actual attack occurs Banking & Finance – Detect and prevent fraud, manage risks, enable personalized services, and automate financial-decision processing Healthcare – Optimize patient interactions, develop personalized treatment plans, attain better patient experience, improve patient data accuracy, and reduce misfiled patient records Manufacturing – Predict maintenance, detect defects and quality issues, enhance productivity, generate product & component designs, and optimize inventory & demand forecasting Retail – Secure self-checkout that helps loss prevention, optimize retail operations & supply chain, and enhance customer experiences Smart cities & IoT – Manage traffic of autonomous vehicles & self-driving, manage energy consumption, optimize water usage, and streamline waste management through real-time sensor data Telecom – Predict network congestion and proactively reroute traffic to avoid outages Bad Cybercriminals – Leverage AI-powered tools and social engineering to steal identities, generate ransomware attacks, perform targeted national state attacks, and destroy national critical infrastructure Computing resources – Require heavy power supply, Thermal Design Power (TDP), Graphics Processing Unit (GPU), and Random Access Memory (RAM) Environmental impact - Impact of intensive computing resources have on carbon footprint and environment Energy cost – Rise in electric power usage and water for cooling and increasing computational costs translates into carbon emissions Bias & Discrimination - Propagate biases as a result of bad training data, incomplete data, and poorly trained AI model Inequality – Widen the gap between the rich and poor and increase inequality in society Privacy – Loss of data privacy from insecure AI systems, unencrypted data sources, and misuse & abuse Skills loss - Reduce human critical thinking skills to uncover root issues, solve complex problems, and ability to write at college level and professionally Scary Job loss and displacement - Replace humans with robots across every sector to perform highly skilled professional jobs Overreliance on AI – Rely heavily on AI to make important decisions like electing medical procedures, making life or death decisions, or choosing political candidates Dominance of AI - Potential ability of AI to surpass human intelligence and take control Monopoly by tech – a select number of tech companies could monopolize the economy and have undue influence over the social construct of our daily lives from buying patterns to everyday decision-making Deepfakes – Generate deepfakes with manipulated videos and images to influence discussions on social media and online forums Propaganda & Disinformation – Deploy human a |
| Envoyé | Oui |
| Condensat | “much 2023 24x7 ability about abuse access according accuracy achieve achieving across actual advancement adversaries against all alone also americans anomalies another application approach are aren’t argue arguments attack attacks attain audited augment authenticity automate automation autonomous available average avoid aware away back bad banking battlefield before behavioral being benefit benefits best better between bias biases body boost bot both but buying came campaigns can candidates capabilities capital carbon care censorship chain challenges change chatbots checkout choices choosing cisco’s cities citizens’ clear climate climbing close college combatting come comes comfortable communicate companies complex component comprehensive computational computing concern concerned concerns congestion connect consequential construct consulting consumer consumers consumption content continues control cooling cost costs could creating creation creativity criminal criminals critical current customer cyber cybercriminals cybersecurity cycles daily data days death decision decisions deepfakes defects defend demand democratic demographics depends deploy design designs despite destroy detect determine detriment detrimental develop developers developing development digital discrimination discussions disinformation displacement doctors does doing dominance don’t doom doomsday doors driving each early economy education efficiency efficiently either electing electric embedding emissions enable energy engine engineering engineers enhance enough ensuring environment environmental era errors ethics every everyday evil evolve example examples excited existed existential expect experience experiences experts explaining explore eyebrows fact familiar far fate fear fetch fiction finance financial finding first footprint forecasting forums forward found fraud freedom from further future gap gdpr gen genai generate generation generative get glean global gloom going good google gpu graduates graphics grew growing grown growth guidance guide have having healthcare hear heavily heavy help helpful helps here higher highly hit holding home hope how human humanity humans hundreds hurt hurtful identities image images impact implications important improve income incomplete increase increased increasing indicated individual individuals industry inequality infinite influence information infrastructure inherent innovations insecure instituting intelligence intelligently intensive interactions interested internet inventory involvement iot isn’t issues it’s its itself job jobs just keep key know knowledge lags law laws learn learning legal let let’s level leverage life lifecycle like little lives look lose loss lot machine machines maintenance major make making manage management manager managing manipulate manipulated manufacturing many market marketer mature maximizing media medical memory mindset minimizing misfiled misuse mitigate mitigates mitigating model modern monopolize monopoly more more” much mundane national navigate navigating necessary need needs negative network never new next not number numerous occurs one online only opens operations opinion optimistic optimization optimize optimizes optimizing organization organizations out outages over overreliance own patient patterns people perceived perform performing personal personalized perspective pessimistic pew places plans points political poor poorly pose positive possibilities potential power powered practice predict predicting pretty prevent prevented prevention previous privacy private proactively problems procedures processes processing product productivity products professional professionally program progress propaganda propagate propel proper protect protection providing public quality raise ram random range ransomware real reality really recommendations records reduce reducing regulate regulating regulation rely removing repetitive replace replaced require reroute research resilience resources respectively respond respondents rest r |
| Tags | Ransomware Tool Prediction Medical |
| Stories | |
| Notes | ★★★ |
| Move | 
|
L'article ne semble pas avoir été repris aprés sa publication.
L'article ne semble pas avoir été repris sur un précédent.