Kopanyo (Alignment)
Alignment
Tiragalo ya go netefatsa gore maikaelelo, diphelelo, le boitshwaro jwa tsamaiso ya AI di tsamaelana le maikaelelo le boleng jwa batho. Se se botlhokwa segolobogolo mo ditsamaisong tse di raraaneng tse di ka tlhabololang boitshwaro jo bo sa ikaelelwang ka tlhamalalo.
Sekai: Go netefatsa gore chatbot ya tshegetso ya boitekanelo jwa tlhaloganyo ga e ke e kgothaletsa ditiro tse di kotsi go sa kgathalesege dikgothatso.
Application Programming Interface (API) (Application Programming Interface (API))
Application Programming Interface (API)
Setlhopha sa melao le diprothokholo tse di tlhalositsweng tse di letlelelang ditsamaiso tse di farologaneng tsa software go buisana le go ananya data.
Sekai: Go dirisa OpenAI API go romela kgothatso le go amogela karabo e e dirilweng ke mokgwa wa puo mo app ya gago ya webo.
Botlhale jwa Maiketsetso jo bo Akaretsang (AGI) (Artificial General Intelligence (AGI))
Artificial General Intelligence (AGI)
Mofuta wa AI wa thiori o o ka dirang tiro nngwe le nngwe ya botlhale e motho a ka e dirang. E akaretsa go ithuta mo mafapheng otlhe.
Sekai: Tsamaiso ya AGI e ka ithuta go kwala mmino, go dira karo, le go atlelela tlhatlhobo ya filosofi ntle le thulaganyo e e rileng ya tiro.
Botlhale jwa Maiketsetso (AI) (Artificial Intelligence (AI))
Artificial Intelligence (AI)
Go etsa botlhale jwa motho mo metšhine e e thulagantsweng go akanya, go tlhaloganya, le go dira ka boikemelo.
Sekai: AI e dirisa bathusi ba botho jaaka Siri le ditsamaiso tsa go kgweetsa ka boikemelo jaaka Tesla Autopilot.
Melao ya Boitshwaro ya AI (AI Ethics)
AI Ethics
Thuto e e amegang ka ditlamorago tsa boitshwaro tsa tlhabololo le tiriso ya AI, go akaretsa toka, boiphitlho, boikarabelo, le go sa kgetholole.
Sekai: Go dira ditaelo go thibela ditsamaiso tsa go hira go kgetholola go ya ka bong kgotsa lotso.
Botlhale jo bo Okeditsweng (Augmented Intelligence)
Augmented Intelligence
Mokgwa wa tšhomišano o AI e tlatsang le go tokafatsa botlhale jwa motho go na le go bo emela.
Sekai: Didirisiwa tsa radiology tse di dirisang AI tse di supang dilo tse di sa tlwaelegang mo dingakeng, tse di dirang tlhatlhobo ya bofelo.
Moemedi yo o Ikemetseng (Autonomous Agent)
Autonomous Agent
Tsamaiso ya AI e e kgonang go dira ditshwetso tsa yona le go tsaya ditiro go fitlhelela maikaelelo a yona ntle le go tsenelela ga motho.
Sekai: Roboto ya go tsamaisa e e ikemetseng e e tsamayang mo ditseleng tsa toropo le go tila dikgoreletsi ka boikemelo.
Backpropagation (Backpropagation)
Backpropagation
Mokgwa wa go thapisa marangrang a neural ka go fetola ditlhotlhwa ka go boela morago go tswa go diphelelo go ya go dikarolo tsa go tsena, go fokotsa diphoso tsa go bolelela pele.
Sekai: E dirisiwa mo go thapiseng dikarolo tsa ditshwantsho go fokotsa sekgala sa diphoso mo go lemogeng dinomoro tse di kwadilweng ka seatla.
Tshekamelo (Tshekamelo ya Algorithmic) (Bias (Algorithmic Bias))
Bias (Algorithmic Bias)
Go rata ka tsela e e sa ikaelelwang le e e tlwaelegileng mo diphelelong tsa AI ka ntlha ya data ya thapiso e e sa lekalekaneng kgotsa e e sa emeleng.
Sekai: Tsamaiso ya go lemoga difatlhego e e sa lemogeng batho ba mmala gantsi ka ntlha ya go sa emelwe sentle mo data ya thapiso.
Data e Kgolo (Big Data)
Big Data
Ditsamaiso tse dikgolo thata tsa data tse di tlhokang didirisiwa tse di kgethegileng go boloka, go sekaseka, le go ntsha boleng, gantsi di dirisiwa go thapisa dimodela tsa AI.
Sekai: Go dirisa dimilione tsa ditsamaisano tsa basebedisi go thapisa dienjine tsa kgothatso tsa dipolatform tsa e-commerce.
Black Box Model (Black Box Model)
Black Box Model
Mofuta wa AI kgotsa mokgwa wa go ithuta ga metšhine o o nang le tlhaloganyo ya ka fa gare e e sa tlhaloganyesegeng motlhofo ke batho, go dira gore go nne thata go tlhaloganya gore ditshwetso di dirwa jang.
Sekai: Marangrang a neural a a botebo a a dirisiwang go amogela dikadimo mme a sa tlhalose sentle gore ke eng fa mokopi mongwe a amogetswe mme yo mongwe a ganetswe.
Go Bala ka Botlhale (Cognitive Computing)
Cognitive Computing
Ditsamaiso tsa AI tse di dirilweng go etsa ditsamaiso tsa go akanya tsa motho, jaaka go akanya le go ithuta, go dirisa mekgwa jaaka NLP le go lemoga dipaterone.
Sekai: Tsamaiso ya go bala ka botlhale e e thusang baitseanape ba molao go sekaseka molao wa kgetsi le go bolelela pele diphelelo.
Pono ya Khomputara (Computer Vision)
Computer Vision
Lefapha la botlhale jwa maiketsetso le le letlelelang dikhomputara go tlhaloganya le go dirisa data ya pono jaaka ditshwantsho le dibidio.
Sekai: Ditsamaiso tsa go lemoga difatlhego tse di lemogang batho mo ditshwantshong tsa tshireletso go dirisa pono ya khomputara.
Corpus (Corpus)
Corpus
Kokoanyo e kgolo ya ditlhakwa tse di kwadilweng kgotsa tse di buiwang tse di dirisiwang go thapisa dimodela tsa puo.
Sekai: Common Crawl dataset ke corpus ya webo ya botlhe e e dirisiwang go thapisa dimodela tse dikgolo tsa puo jaaka GPT.
Data Drift (Data Drift)
Data Drift
Tiragalo e data ya go tsena e fetogang ka nako, e dira gore tiriso ya mokgwa e fokotsege.
Sekai: Mokgwa wa tlhokomelo ya go bolelela pele mo didirisiweng tsa indasteri o nna o sa nepahala fa thekenoloji e ntšha ya sensor e tsenngwa.
Data Labelling (Data Labelling)
Data Labelling
Tiragalo ya go kwala data ka ditshwao kgotsa dileibole go e dira gore e siamele go ithuta ka tlhokomelo.
Sekai: Go kwala dikete tsa ditshwantsho tsa kankere jaaka tse di sa kotsi kgotsa tse di kotsi go thapisa mokgwa wa go lemoga kankere.
Data Mining (Data Mining)
Data Mining
Tiragalo ya go lemoga dipaterone tse di nang le bokao, dikamano, le dilo tse di sa tlwaelegang mo ditsamaisong tse dikgolo tsa data.
Sekai: Barekisi ba dirisa data mining go lemoga gore batho ba ba rekang diaparo tsa masea gantsi ba reka le bojalwa.
Go Ithuta ka Botebo (Deep Learning)
Deep Learning
Lefapha le lennye la go ithuta ga metšhine le le dirisang marangrang a neural a a nang le dikarolo tse dintsi go etsa dipaterone tse di raraaneng mo data.
Sekai: Go ithuta ka botebo go dirisiwa mo dimodeleng tsa puo jaaka GPT-4 le dimodela tsa go dira ditshwantsho jaaka Stable Diffusion.
Diffusion Models (Diffusion Models)
Diffusion Models
Setlhopha sa dimodela tsa go dira tse di ithutang go dira data ka go fetola modumo o o sa tlwaelegang go nna diphelelo tse di rulagantsweng.
Sekai: Stable Diffusion e dira ditshwantsho tsa photorealistic go tswa go dikgothatso tsa ditlhakwa go dirisa mekgwa ya diffusion.
Embedding (Embedding)
Embedding
Kemedi ya nomoro ya data, gantsi e dirisiwa go tshwara bokao jwa semantic jwa mafoko, ditshwantsho, kgotsa dipolelo.
Sekai: Mo NLP, lefoko 'bank' le ka nna la nna le di-embedding tse di tshwanang le 'money' mme di farologane le 'riverbank' go ya ka seemo.
Epoch (Epoch)
Epoch
Tiragalo e e feletseng ya go feta mo data ya thapiso yotlhe ka nako ya tiragalo ya thapiso ya mokgwa wa go ithuta ga metšhine.
Sekai: Fa data e na le dikai di le 1,000 mme mokgwa o di bona tsotlhe gangwe ka nako ya thapiso, ke epoch e le nngwe.
Ethical AI (Ethical AI)
Ethical AI
Filosofi ya go rala le go dirisa e e netefatsang gore dithekenoloji tsa AI di dira ka phepafalo, ka toka, le go ya ka boleng jwa setšhaba.
Sekai: Sedirisiwa sa go hira sa AI se se akaretsang ditlhatlhobo tsa tshekamelo go thibela kgethololo kgatlhanong le bakopi ba ba tswang mo ditlhopheng tse dinnye.
Expert System (Expert System)
Expert System
Tsamaiso ya AI e e etsang bokgoni jwa go dira ditshwetso jwa moitseanape wa motho mo lefelong le le rileng go dirisa melao le tlhaloganyo.
Sekai: Tsamaiso ya moitseanape e e dirisiwang mo temothuong go kgothaletsa ditlhabololo tsa dijalo go ya ka data ya mmu le hisitori ya disenyi.
Explainable AI (XAI) (Explainable AI (XAI))
Explainable AI (XAI)
Ditsamaiso tsa AI tse di dirilweng go dira gore ditsamaiso tsa tsona tsa ka fa gare le ditshwetso di tlhaloganyesegwe ke batho, go oketsa tshepo le boikarabelo.
Sekai: AI ya tlhatlhobo ya bongaka e e sa kgothaletseng fela mme gape e tlhalosa gore ke dife ditshupo tse di lebisitseng kwa tlhagiso eo.
Few-shot Learning (Few-shot Learning)
Few-shot Learning
Mokgwa wa go ithuta ga metšhine o mokgwa o thapisitsweng kgotsa o fetotsweng go dirisa dikai di le mmalwa fela tse di kwadilweng.
Sekai: Go fetola LLM go kwala di-imeile tsa molao morago ga go e bontsha dikai di le 10 fela.
Fine-tuning (Fine-tuning)
Fine-tuning
Tiragalo ya go tsaya mokgwa o o thapisitsweng pele le go o thapisa gape mo data e ntšha, e nnye go o kgethegisa mo tirong e e rileng.
Sekai: Go fetola LLM e e tlwaelegileng jaaka GPT mo ditlhakwaneng tsa molao tsa ka fa gare go dira mothusi wa go kwala molao.
Foundation Model (Foundation Model)
Foundation Model
Mokgwa o mogolo o o thapisitsweng mo data e e farologaneng le e e pharaletseng e e ka fetolwang go ditiro tse dintsi tsa kwa tlase.
Sekai: GPT-4 le PaLM 2 ke dimodela tsa motheo tse di kgonang go akaretsa, go botsa le go araba dipotso, go ranola, le tse dingwe.
Fuzzy Logic (Fuzzy Logic)
Fuzzy Logic
Mofuta wa tlhaloganyo o o dirisanang le boleng jo bo batlang go nna jalo go na le tlhaloganyo e e tlhomameng ya nnete/maaka (binary), e e mosola mo go akanyeng ka fa tlase ga go sa tlhomamisege.
Sekai: E dirisiwa mo ditsamaisong tsa go laola themperetšha go fetola themperetšha go ya ka dikarolo tsa fuzzy jaaka 'go mogote go sekae' kgotsa 'go tsididi thata'.
Generative Adversarial Network (GAN) (Generative Adversarial Network (GAN))
Generative Adversarial Network (GAN)
Mokgwa wa go dira o marangrang a mabedi — generator le discriminator — a gaisanang go tokafatsa boleng jwa diphelelo.
Sekai: GANs di dirisiwa go dira dibidio tsa deepfake kgotsa go dira ditshwantsho tsa dikgwebo tse di lebegang di le tsa nnete go tswa go disketšhe.
Generative AI (Generative AI)
Generative AI
Setlhopha sa botlhale jwa maiketsetso se se ka dirang dikgatiso tse disha — jaaka ditlhakwa, ditshwantsho, mmino, kgotsa dibidio — go tswa go data ya thapiso.
Sekai: ChatGPT e dira diposo tsa blog kgotsa Midjourney e dira ditshwantsho tsa dijithale go tswa go dikgothatso tsa ditlhakwa.
Generative Pre-trained Transformer (GPT) (Generative Pre-trained Transformer (GPT))
Generative Pre-trained Transformer (GPT)
Setlhopha sa dimodela tse dikgolo tsa puo tse di tlhabolotsweng ke OpenAI tse di dirisang transformer architecture mme di thapisitswe pele mo ditsamaisong tse dikgolo tsa data ya ditlhakwa go dira ditiro tse di farologaneng tsa puo.
Sekai: GPT-4 e kgona go kwala di-essay, go ranola dipuo, le go akaretsa ditlhakwa ka kgothatso e nnye.
Genetic Algorithm (Genetic Algorithm)
Genetic Algorithm
Mokgwa wa go tokafatsa o o tlhotlheleditsweng ke tlhopho ya tlholego o ditlhabololo di tlhabologang ka nako ka go fetoga, go tswakana, le tlhopho.
Sekai: E dirisiwa go rala marangrang a neural a a nang le bokgoni ka go etsa go falola ga ba ba nonofileng.
Hallucination (Hallucination)
Hallucination
Go dira dikgatiso tse di lebegang di le tsa nnete mme di sa nepahala kgotsa di se na tlhaloganyo ke mokgwa wa AI.
Sekai: Mokgwa wa puo o tlhama tlhagiso e e seng teng kgotsa o fana ka dintlha tsa hisitori tse di sa nepahalang.
Heuristic (Heuristic)
Heuristic
Mokgwa o o dirisiwang go rarabolola mathata o o sa netefatseng tharabololo e e feletseng mme o lekane mo maikaelelong a gompieno.
Sekai: Go dirisa molao wa tlwaelo go lekanyetsa nako ya go tsamaisa mo tsamaisong ya AI ya logistics.
Hyperparameter (Hyperparameter)
Hyperparameter
Boleng jwa phetolo bo bo tlhomilweng pele ga go thapisa mokgwa wa go ithuta ga metšhine, jaaka sekgala sa go ithuta kgotsa palo ya dikarolo.
Sekai: Go fetola batch size go tswa go 32 go ya go 128 go tokafatsa lebelo la thapiso le tiriso ya mokgwa.
Inference (Inference)
Inference
Tiragalo ya go dirisa mokgwa wa go ithuta ga metšhine o o thapisitsweng go dira dipolelelo kgotsa go dira diphelelo go tswa go data e ntšha ya go tsena.
Sekai: Go dirisa mokgwa wa GPT o o fetotsweng go kwala di-imeile tsa setlhopha sa tshegetso ya bareki.
Intent Detection (Intent Detection)
Intent Detection
Tiro mo go tlhaloganyeng puo ya tlholego e tsamaiso e lemogang maikaelelo kgotsa boikaelelo jwa mosebedisi mo molaetseng.
Sekai: Mo chatbot, go lemoga 'Ke batla go booka sefofane' jaaka maikaelelo a go booka maeto.
Internet of Things (IoT) (Internet of Things (IoT))
Internet of Things (IoT)
Marangrang a didirisiwa tsa mmele tse di golaganeng tse di nang le disensor, software, le dithekenoloji tse dingwe go kokoanya le go ananya data.
Sekai: Dithemperetšha tse di botlhale le difridge tse di begang data ya tiriso le go fetola ditlhophiso go dirisa AI analytics.
Interpretability (Interpretability)
Interpretability
Tekanyetso e motho a ka tlhaloganyang ka yona ditsamaiso tsa ka fa gare tsa mokgwa wa go ithuta ga metšhine le tiragalo ya ona ya go dira ditshwetso.
Sekai: Sefate sa ditshwetso se tlhaloganyesega go feta marangrang a neural a a botebo ka gonne ditshwetso tsa sona di ka latelwa.
Jupyter Notebook (Jupyter Notebook)
Jupyter Notebook
Tikologo ya go bala e e dirisanang e e bulegileng e e letlelelang basebedisi go kwala khoutu, go bona diphelelo, le go kwala tlhatlhobo mo segokaganyong se le sengwe.
Sekai: Baitseanape ba data ba dirisa Jupyter Notebooks go dira dimodela tsa go ithuta ga metšhine le go abelana diphelelo.
K-Nearest Neighbours (KNN) (K-Nearest Neighbours (KNN))
K-Nearest Neighbours (KNN)
Mokgwa o o motlhofo, o o sa tlwaelegang wa go ithuta ga metšhine o o dirisiwang mo go kgethololeng le go tlhatlhoba. O dira ditshwetso go ya ka dikai tsa thapiso tse di gaufi thata mo lefelong la dintlha.
Sekai: Go kgetholola leungo le lesha jaaka apole kgotsa pere, KNN e tlhatlhoba gore ke dife maungo a a kwadilweng a a gaufi thata ka sebÅpego le mmala.
Knowledge Graph (Knowledge Graph)
Knowledge Graph
Sebopego sa data se se dirisang dinode le ditsela go emela le go boloka ditlhaloso tse di golaganeng tsa dilo le dikamano tsa tsona.
Sekai: Panel ya kitso ya Google e dirisiwa ke knowledge graph e e golaganyang dilo jaaka batho, mafelo, le ditiragalo.
Language Learning Model Optimisation (LLMO) (Language Learning Model Optimisation (LLMO))
Language Learning Model Optimisation (LLMO)
Mekgwa e e dirisiwang go tokafatsa tiriso, bokgoni, kgotsa go fetoga ga dimodela tse dikgolo tsa puo mo ditiro kgotsa mafelo a a rileng.
Sekai: Go dirisa quantisation le instruction tuning go tokafatsa LLM mo tirisong ya kgwebo.
Large Language Model (LLM) (Large Language Model (LLM))
Large Language Model (LLM)
Mofuta wa mokgwa wa go ithuta ka botebo o o thapisitsweng mo ditsamaisong tse dikgolo tsa data ya ditlhakwa o o kgonang go dira, go tlhaloganya, le go akanya ka puo ya motho.
Sekai: ChatGPT le Claude ke LLMs tse di thapisitsweng go thusa mo go kwala, go khouta, le go araba dipotso.
Latent Space (Latent Space)
Latent Space
Kemedi e e raraaneng ya data e e nang le dikarolo tse dintsi e dikarolo tse di tshwanang di kokoanngwang gaufi, e e dirisiwang mo dimodeleng tsa go dira le di-embedding.
Sekai: Mo go diriseng ditshwantsho, go dirisa latent space go ka fetola dintlha jaaka go phatsima kgotsa maikutlo.
Learning Rate (Learning Rate)
Learning Rate
Hyperparameter ya botlhokwa mo thapisong e e laolang gore ditlhotlhwa tsa mokgwa di fetolwa go le kae go ya ka loss gradient.
Sekai: Sekgala se segolo sa go ithuta se ka lebisa kwa go feteng minima, fa sekgala se se kwa tlase thata se fokotsa lebelo la thapiso.
Machine Learning (ML) (Machine Learning (ML))
Machine Learning (ML)
Lefapha la AI le le letlelelang ditsamaiso go ithuta go tswa go data le go tokafatsa tiriso ntle le go thulaganyetswa ka tlhamalalo.
Sekai: Difilthara tsa spam di dirisa go ithuta ga metšhine go kgetholola di-imeile jaaka spam kgotsa eseng go ya ka dikai tsa nako e e fetileng.
Model Drift (Model Drift)
Model Drift
Tiragalo e tiriso ya mokgwa e fokotsegang ka nako ka ntlha ya diphetogo mo data kgotsa tikologong.
Sekai: Mokgwa wa go lemoga boferefere o nna o sa nepahala fa mekgwa ya boferefere e tlhabologa.
Model Training (Model Training)
Model Training
Tiragalo ya go fepa data go mokgwa wa go ithuta ga metšhine le go fetola diparameter tsa ona go fokotsa phoso.
Sekai: Go thapisa engine ya kgothatso mo hisitoring ya theko ya bareki go kgothaletsa dikgwebo tse disha.
Multimodal AI (Multimodal AI)
Multimodal AI
Ditsamaiso tsa AI tse di kgonang go dirisa le go kopanya mefuta e le mentsi ya data jaaka ditlhakwa, ditshwantsho, modumo, le dibidio.
Sekai: Mokgwa jaaka GPT-4 Vision o o kgonang go bala ditlhakwa le go tlhaloganya ditshwantsho ka nako e le nngwe.
Natural Language Processing (NLP) (Natural Language Processing (NLP))
Natural Language Processing (NLP)
Lefapha le lennye la AI le le tsepamisitseng mogopolo mo tirisanong magareng ga dikhomputara le dipuo tsa batho (tlholego). Le letlelelang metšhine go bala, go tlhaloganya, le go araba ka puo ya motho.
Sekai: NLP e dirisiwa mo bathusing ba lentswe, di-app tsa go ranola puo, le di-chatbot.
Neural Network (Neural Network)
Neural Network
Mokgwa wa go ithuta ga metšhine o o tlhotlheleditsweng ke sebopego sa boboko jwa motho, o o dirilweng ka dikarolo tsa dinode tse di golaganeng (neurons).
Sekai: Marangrang a neural a kwa morago ga dimodela tsa go ithuta ka botebo tse di dirisiwang mo go lemogeng ditshwantsho le puo.
Noise (Noise)
Noise
Tshedimosetso e e sa tlwaelegang kgotsa e e sa amaneng mo data e e ka fitlhang dipaterone tse di nang le bokao le go ama tiriso ya mokgwa ka tsela e e sa siamang.
Sekai: Diphoso tsa sensor kgotsa dintlha tsa data tse di tletseng diphoso di ka tsewa jaaka modumo.
Ontology (Ontology)
Ontology
Sebopego se se rulagantsweng se se kgethololang le go tlhalosa dikamano magareng ga dikgopolo mo lefelong le le rileng, gantsi e dirisiwa mo ditsamaisong tsa semantic AI.
Sekai: Ontology mo boitekanelong e ka tlhalosa gore ditshupo di amana jang le malwetse le ditlhabololo.
Overfitting (Overfitting)
Overfitting
Phoso ya go dira mokgwa o mokgwa wa go ithuta ga metšhine o tshwarang modumo mo data ya thapiso mme o dira bobe mo data e ntšha.
Sekai: Mokgwa o o gopolang dikarabo tsa thapiso mme o sa kgone go dirisa data ya tlhatlhobo e e sa bonweng o overfitted.
Predictive Analytics (Predictive Analytics)
Predictive Analytics
Tiriso ya data, algorithms, le AI go lemoga kgonagalo ya diphelelo tsa nako e e tlang go ya ka data ya hisitori.
Sekai: Barekisi ba dirisa predictive analytics go bolelela pele tlhokego ya dikgwebo tse di rileng.
Pre-training (Pre-training)
Pre-training
Tiragalo ya go thapisa mokgwa pele mo data e kgolo, e e tlwaelegileng pele ga go o fetola mo ditiro tse di rileng.
Sekai: Dimodela tsa GPT di thapisitswe pele mo ditsamaisong tse dikgolo pele ga go di fetola mo di-chatbot tsa tshegetso ya bareki.
Prompt Engineering (Prompt Engineering)
Prompt Engineering
Botaki le saense ya go dira dikgothatso tse di nang le bokgoni go laola diphelelo tsa dimodela tse dikgolo tsa puo.
Sekai: Go oketsa ditaelo tsa tsamaiso jaaka 'Araba jaaka morutabana yo o maitseo' ke sekai sa prompt engineering.
Quantisation (Quantisation)
Quantisation
Mokgwa wa go gatelela mokgwa o o fokotsang palo ya dibiti tse di dirisiwang go emela ditlhotlhwa le ditiro, go tokafatsa bokgoni.
Sekai: Go fetola mokgwa go tswa go 32-bit go ya go 8-bit go tokafatsa tiriso mo didirisiweng tsa selefouno.
Quantum Computing (Quantum Computing)
Quantum Computing
Mokgwa o mosha wa go bala o o theilweng mo quantum mechanics, o o nang le bokgoni jwa go dira ditsamaiso ka lebelo le le kwa godimo.
Sekai: Quantum computing e ka nna ya tokafatsa thapiso ya AI go feta ditekanyetso tsa tlwaelo.
Reasoning Engine (Reasoning Engine)
Reasoning Engine
Tsamaiso mo AI e e ntshang ditshwetso tsa tlhaloganyo go tswa go setlhopha sa dintlha kgotsa data go dirisa melao kgotsa algorithms ya inference.
Sekai: Sedirisiwa sa tlhatlhobo ya AI se se dirisang reasoning engine go ntsha maemo a bongaka a a kgonagalang go ya ka ditshupo.
Reinforcement Learning (RL) (Reinforcement Learning (RL))
Reinforcement Learning (RL)
Lefapha la go ithuta ga metšhine le baemedi ba ithutang ka go dirisana le tikologo ya bona go oketsa dituelo tse di kokoantsweng.
Sekai: Roboto e e ithutang go tsamaya ka go leka le go dira diphoso go dirisa mekgwa ya RL.
Reinforcement Learning with Human Feedback (RLHF) (Reinforcement Learning with Human Feedback (RLHF))
Reinforcement Learning with Human Feedback (RLHF)
Mokgwa wa go ithuta o dikgothatso tsa batho di laolang letswao la tuelo la AI, gantsi e dirisiwa mo go fetoleng dimodela tsa puo.
Sekai: ChatGPT e thapisitswe ka RLHF go dira dikarabo tse di mosola le tse di sireletsegileng.
Retrieval-Augmented Generation (RAG) (Retrieval-Augmented Generation (RAG))
Retrieval-Augmented Generation (RAG)
Mokgwa o o kopanyang go batla tshedimosetso le go dira, o LLM e batlang ditlhakwa tse di amanang go tokafatsa karabo ya yona.
Sekai: Mothusi wa AI o batla le go nopola dintlha tsa kgwebo fa a dira karabo ya potso ya botegeniki.
Self-Supervised Learning (Self-Supervised Learning)
Self-Supervised Learning
Mokgwa wa thapiso o mokgwa o ithutang dipaterone ka go dira dileibole tsa ona go tswa go data e e sa dirisiwang, go fokotsa go ikaega ka data e e kwadilweng ke motho.
Sekai: BERT e thapisitswe ka self-supervised learning ka go bolelela pele mafoko a a tlhokang mo ditlhakwaneng.
Semantic Search (Semantic Search)
Semantic Search
Mokgwa wa go batla o o tlhaloganyang maikaelelo a mosebedisi le bokao jwa seemo, eseng fela go tshwanelana le mafoko a botlhokwa.
Sekai: Go batla 'mokgwa wa go lokisa pompo e e dutlang' go busetsa ditaelo le fa lefoko 'pompo e e dutlang' le seyo mo tlhakwaneng.
Sentiment Analysis (Sentiment Analysis)
Sentiment Analysis
Tiragalo ya go lemoga maikutlo, dikgopolo, kgotsa maikutlo mo ditlhakwaneng, gantsi e kgetholola jaaka e e siameng, e e sa siamang, kgotsa e e sa kgethololeng.
Sekai: Go sekaseka di-tweet go lekanyetsa karabo ya botlhe mo kgwebong e ntšha.
Stochastic (Stochastic)
Stochastic
Go akaretsa go sa tlwaelegang kgotsa boitshwaro jwa go kgonagala, gantsi e dirisiwa mo go diriseng AI le algorithms ya go tokafatsa.
Sekai: Diphelelo tsa GPT-4 di farologana mo go tsengweng go go tshwanang ka ntlha ya tiragalo ya yona ya stochastic decoding.
Strong AI (Strong AI)
Strong AI
Gape e itsege jaaka Botlhale jwa Maiketsetso jo bo Akaretsang (AGI), e bua ka metšhine e e nang le bokgoni jwa botlhale jwa motho mo mafapheng otlhe.
Sekai: AI ya nako e e tlang e e ka kwalang dinovel, go rala ditoropo, le go rarabolola mathata a boitshwaro ka go tshwana.
Super Artificial Intelligence (SAI) (Super Artificial Intelligence (SAI))
Super Artificial Intelligence (SAI)
AI ya thiori e e fetang botlhale jwa motho mo dintlheng tsotlhe—go akanya, bokgoni, botlhale jwa maikutlo, jalo le jalo.
Sekai: SAI e ka nna ya tlhabolola disayense le difilosofi tse disha ka boikemelo.
Supervised Learning (Supervised Learning)
Supervised Learning
Mokgwa wa go ithuta ga metšhine o dimodela di thapisitsweng mo data e e kwadilweng go ithuta go golaganya go tsena le go tswa.
Sekai: Go ruta mokgwa go kgetholola di-imeile jaaka spam kgotsa eseng go dirisa dikai tsa hisitori.
Synthetic Data (Synthetic Data)
Synthetic Data
Data e e dirilweng ka maiketsetso e e etsang data ya nnete, gantsi e dirisiwa mo thapisong fa data ya nnete e le nnye kgotsa e le botlhokwa.
Sekai: Go dira ditshwantsho tsa bongaka tsa synthetic go thapisa dimodela tsa tlhatlhobo ntle le go tlola boiphitlho jwa molwetse.
Token (Token)
Token
Yuniti ya ditlhakwa e e dirisiwang ke LLMs—gantsi lefoko kgotsa karolo ya lefoko.
Sekai: Polelo 'Lefatshe!' e kgaogantswe ka ditokene di le 3: 'Lefatshe', '!', le '!'.
Tokenisation (Tokenisation)
Tokenisation
Tiragalo ya go kgaoganya ditlhakwa go nna ditokene go dirisiwa ke mokgwa.
Sekai: Mo NLP, 'ChatGPT e ntle' e nna ['Chat', 'G', 'PT', 'e', 'ntle'].
Transfer Learning (Transfer Learning)
Transfer Learning
Go dirisa kitso go tswa go tiro e le nngwe go tokafatsa go ithuta mo tirong e nngwe e e amanang, go fokotsa nako ya thapiso le ditlhokego tsa data.
Sekai: Go fetola mokgwa o o thapisitsweng mo ditlhakwaneng tsa Seesemane go dira tlhatlhobo ya maikutlo mo puong e nngwe.
Transformer (Transformer)
Transformer
Marangrang a neural architecture a a dirisang mekgwa ya tlhokomelo go etsa data ya tatelano, e e dirisiwang thata mo LLMs.
Sekai: BERT, GPT, le T5 ke dimodela tse di theilweng mo transformer.
Underfitting (Underfitting)
Underfitting
Fa mokgwa o le motlhofo thata go tshwara dipaterone mo data ya thapiso, go dira gore tiriso e nne bobe.
Sekai: Mokgwa wa linear o o lekang go bolelela pele dikgethololo tsa ditshwantsho tse di raraaneng o ka nna wa underfit.
Unsupervised Learning (Unsupervised Learning)
Unsupervised Learning
Mokgwa wa go ithuta o dimodela di lemogang dipaterone kgotsa ditlhopha mo data e e sa kwadilweng.
Sekai: Go tlhopha bareki go ya ka boitshwaro jwa theko ntle le dileibole tse di tlhalositsweng pele.
User Intent (User Intent)
User Intent
Maikaelelo kgotsa boikaelelo jwa potso kgotsa tirisano ya mosebedisi.
Sekai: Mosebedisi yo o kwalang 'mokgwa wa go baka kuku' o ka nna a ikaelela go batla risepe.
Validation Set (Validation Set)
Validation Set
Karolo ya data e e dirisiwang go tlhatlhoba tiriso ya mokgwa ka nako ya thapiso le go fetola hyperparameters.
Sekai: E dirisiwa go lemoga overfitting pele ga tlhatlhobo ya bofelo.
Vector Database (Vector Database)
Vector Database
Database e e dirilweng go boloka le go batla vector embeddings tse di dirisiwang mo ditiro tsa AI jaaka similarity search le RAG.
Sekai: Pinecone le Weaviate ke vector databases tsa go boloka ditlhakwa kgotsa ditshwantsho embeddings.
Vector Embedding (Vector Embedding)
Vector Embedding
Kemedi ya nomoro ya data e e bolokang bokao jwa semantic le dikamano mo lefelong la vector.
Sekai: Mafoko 'kgosi' le 'kgosigadi' a na le di-embedding tse di tshwanang le dipharologano tse dinnye tsa bong.
Virtual Assistant (Virtual Assistant)
Virtual Assistant
Moemedi wa software o o dirisang AI o o thusang basebedisi go fetsa ditiro ka puisano kgotsa ditaelo tsa lentswe.
Sekai: Siri, Alexa, le Google Assistant ke bathusi ba botho ba ba tumileng.
Voice Recognition (Voice Recognition)
Voice Recognition
Thekenoloji e e tlhaloganyang le go fetola puo e e buiwang go nna ditlhakwa kgotsa tiro.
Sekai: Go kwala ka lentswe le ditaelo tsa lentswe di ikaegile ka ditsamaiso tsa go lemoga lentswe.
Weak AI (Weak AI)
Weak AI
Ditsamaiso tsa AI tse di dirilweng go dira tiro e e nnye, e e rileng ntle le botlhale jo bo akaretsang.
Sekai: AI ya go tshameka chess e e sa kgoneng go tlhaloganya puo kgotsa go kgweetsa koloi ke sekai sa weak AI.
Web Scraping (Web Scraping)
Web Scraping
Go ntsha tshedimosetso ka boikemelo go tswa go diwebosaete, gantsi e dirisiwa go kokoanya data ya thapiso kgotsa go tlhokomela dikgatiso.
Sekai: Go batla mananeo a matlo go thapisa mokgwa wa go lekanyetsa boleng jwa thoto.
Weight (Weight)
Weight
Parameter mo marangrang a neural e e laolang maatla a tlhotlheletso e node e le nngwe e nang le yona mo go e nngwe.
Sekai: Ditlhotlhwa di fetoga ka nako ya thapiso go fokotsa phoso ya mokgwa.
Whisper (Whisper)
Whisper
Mokgwa wa puo go ya go ditlhakwa o o tlhabolotsweng ke OpenAI o o kgonang go kwala modumo mo dipuong tse dintsi.
Sekai: Whisper e ka kwala dipuo le dipodcast ka nepahalo e kgolo.
YAML (YAML)
YAML
Mokgwa o o balegang ke motho wa go dira data serialisation, o o dirisiwang gantsi mo difaeleng tsa phetolo mo ditsamaisong tsa go ithuta ga metšhine.
Sekai: Go tlhalosa diparameter tsa mokgwa mo faeleng ya YAML go thapisa mo PyTorch.
Zero-shot Learning (Zero-shot Learning)
Zero-shot Learning
Bokgoni jwa mokgwa go dira ditiro tse o sa thapisitsweng ka tlhamalalo mo go tsona ka go dirisa kitso e e tlwaelegileng.
Sekai: Mokgwa o o arabang dipotso tsa molao le fa o sa thapisitswe ka tlhamalalo mo data ya molao.
Zettabyte (Zettabyte)
Zettabyte
Yuniti ya data ya dijithale e e lekanang le sextillion e le nngwe (10^21) ya dibyte, gantsi e dirisiwa go tlhalosa bogolo jwa data ya inthanete.
Sekai: Tiriso ya inthanete ya lefatshe e fetile 1 zettabyte ka ngwaga ka 2016.
