Filter
Exclude
Time range
-
Near
Yes, in one sense Google is declining. In another, Google will rise again. They have a good, diverse portfolio of products & data :) Falling search share due to AI competition. Internal chaos - waves of layoffs, endless reorganizations. Product failures from Stadia to the Gemini controversies. Antitrust lawsuits costing them both money and reputation... They go from mistake to mistake. Like some people we know. :) But - India is saving them. Gemini dominates in coding. It's still a serious company. Maybe rolling downhill, but their future is guaranteed. #Google #Gemini #AI #BigTech #SearchEngine #Antitrust #TechNews #India #GoogleAI #FutureOfTech
13
Le retour aux sources numériques 🔍 @DuckDuckGo Pourquoi la navigation intuitive devient-elle un luxe dans un océan de réponses synthétiques automatisées ? #KingLand #IA #TechTrends #WebSearch #DigitalStrategy #Innovation #UXDesign #FutureOfSearch #MarketingDigital #Algorithm #InformationAge #TechNews #UserExperience #DataManagement #SearchEngine ▫️ Fiche Impact : kingland.fr/article/moteur-d… Parce qu'avec une intention claire, une recherche précise vaut mieux qu'une réponse générée au hasard. L'année 2026 marque un tournant. Après l'euphorie des modèles prédictifs, nous redécouvrons la valeur brute de la recherche textuelle pure. Le besoin de vérité vérifiable, débarrassée des hallucinations algorithmiques, replace le moteur de recherche traditionnel au centre de nos stratégies d'acquisition. Ce n'est pas un retour en arrière, c'est une exigence de transparence retrouvée. 📍 La fiabilité au premier coup d'œil : les sources primaires reprennent leur place dans le classement. 🌐 La liberté d'exploration : retrouver la sérendipité, cette faculté de découvrir des idées que l'IA ne juge pas pertinentes pour vous. 📉 Le contrôle des coûts : réduire la dépendance aux infrastructures gourmandes pour favoriser la légèreté de navigation. ⚖️ L'éthique des données : garantir que le contenu original est valorisé plutôt que simplement digéré par un robot. ✨ Explorer : kingland.fr/tool/duck-ai Le futur de la recherche ne sera pas forcément une intelligence qui pense à notre place, mais une interface capable de nous donner, enfin, les outils pour penser par nous-mêmes. — C. Pestel Je me souviens de l'époque où chercher une information était une aventure intellectuelle. Aujourd'hui, je crains que la facilité offerte par l'IA ne soit en train d'atrophier notre esprit critique. Il est temps de remettre l'outil à sa place : au service de notre curiosité, et non à sa place. Quel outil privilégiez-vous aujourd'hui pour garantir la fiabilité de vos recherches quotidiennes ?
2
19
Mail, Messaging, Browser, Cloud, Socials… the web is full of alternatives to Big Tech which keeps you locked into their ecosystem. Do One Thing Today Try an alternative. It might seem a hassle. It’ll be different. But that’s the point. Independent tech is different and diversity brings choice. mojeek.com #mojeek #search #searchengine #bigtech #alternatives
1
1
5
85
Apple's Biggest Privacy Lie Exposed! Apple markets itself as a privacy-focused company, but there's a side of the story many users don't know. #Apple #Google #Privacy #Technology #Business #Safari #SearchEngine #BigTech #CyberSecurity #DataPrivacy #TechNews #BusinessCaseStudy
12
看完这张全球流量图算是明白了,现在 AI 喊得再响,Google 依然是断层第一。实操下来也是这样,日常最高频的动作还是得靠 Google 搜索。界面干净、没广告,也是大家最容易适应的工具。AI 效率确实高,但真正的“底层基建”还是 Google 啊。 ​​#Google #ChatGPT #Tech #SearchEngine
1
58
Searching online shouldn't feel like work.😏 Hubite.com is a modern search engine built to help users discover profiles, content, and communities faster. Clean results. Better discovery. Less noise. Whether you're looking for new creators, niche interests, or trending profiles, Hubite helps you find what matters without endless scrolling. Powered by CreatorTraffic. Visit Hubite.com and see what you've been missing. 👌 #Hubite #SearchEngine #OnlineSearch #Discovery #CreatorTraffic
33
Three days ago ZENITH had 0.812 Recall@10. I kept debugging. Now it’s 0.906. Key fixes: 1. Weighted RRF (semantic 2x) - fusion now beats dense-only 2. Inverted posting lists -> BM25 from 284ms to 10ms 3. Indexing from 5,536s to 890s (padding batching fixes) ZENITH Hybrid now 45% better recall than SQLite FTS5. Still zero infrastructure. Still one go get. Full writeup: github.com/shramanb113/ZENIT… Would you use this in production? Honest feedback welcome 👇 #Golang #GoLang #Backend #VectorSearch #SearchEngine #OSS @golang @golangch @golangprojects @GolangTrends @golangweekly @GolangRepos
1
3
64
Every time you type: "how to ma..." Google somehow knows you probably mean: "how to make pancakes" before you've finished typing. And it does this in a few milliseconds. That's what makes search autocomplete one of the most fascinating system design problems. At first glance it sounds simple: User types a prefix. Return matching queries. Done. But at Google scale, you're dealing with: • billions of searches • millions of unique queries • thousands of requests per second • strict latency requirements • constantly changing trends The challenge isn't finding matches. The challenge is finding the BEST matches instantly. For example: "ja" could match: • javascript • java • japan • jaguar • java tutorial • java interview questions The system must rank suggestions based on: → popularity → recency → location → personalization → trending events all while responding fast enough to feel instantaneous. The core architecture is surprisingly elegant: → Queries are stored in a Trie (prefix tree) → Each node stores the top-k most popular completions → As the user types, traversal is O(length of prefix) → Results are served from memory, not disk This is why autocomplete feels instant. The real-world challenges are where it gets interesting: • Cache hot prefixes ("a", "how", "best") • Handle breaking news and trending topics • Correct typos and misspellings • Support multiple languages • Prevent abusive or harmful suggestions • Keep rankings fresh without rebuilding everything One viral event can completely change the ranking of a query within minutes. A good autocomplete system isn't just a data structure problem. It's a ranking problem. A caching problem. A distributed systems problem. The chapter covers: • Trie internals • Top-k ranking • Caching strategies • Real-time updates • Sharding • Personalization • Query popularity tracking • Scaling to billions of searches One of the best examples of how a simple user experience hides an incredibly sophisticated backend. Read the full deep dive: hld.handbook.academy/curricu… #SystemDesign #DistributedSystems #BackendEngineering #SoftwareEngineering #SearchEngine
1
40
How Google Secretly Influence Elections A concept known as the **Search Engine Manipulation Effect (SEME)** suggests that simply changing the ranking of search results can significantly influence undecided voters. #Google #SearchEngine #SEME #CyberSecurity #Technology
22
Provenworker retweeted
Only an app goes directly to a website. #Searchengine websites are inferior to predecessor versions that offered the option of typing website address and going directly there. #Search
1
5
Google control about 90% of searches. Microsoft around 5%. Bing proxies fight over the rest. Except Mojeek. 100% independent of Big Tech for free and unbiased access to the wealth of human knowledge on the web. mojeek.com/ #mojeek #search #searchengine #bigtech #alternatives
1
4
110
I’m not Anti-Apple, I like them Google case is something way bigger, not only about their personal assistant, mostly about their searchEngine and favoring their products. With my whole dislike for them, at least I can talk in my native language to their assistant or simply use it
11
🚀 SEO is a marathon, not a sprint! Build lasting rankings, attract quality traffic, and grow your business with proven strategies from Sieve Softech.#SieveSoftech #SEO #SEOServices #GoogleRanking #DigitalMarketing #OrganicTraffic #LeadGeneration #BusinessGrowth #SearchEngine
4