Seharusnya hari ini proses hari ke 110. Namun sekitar 3 harian tidak mood luar biasa untuk belajar.

Hari ini saya memutuskan untuk mengambil course yang lain dulu. Beralih dari foundry ke course Introduction to Artificial Intelligence.

Ini gara-gara ada pengumuman dari Himmata UT. Ada course dari dicoding. Course seribu umat. Iseng-iseng saya cobalah. Ternyata membutuhkan requirement dari course IBM.

Alhamdulillah beres coursenya, dengan catatan-catatan yang saya ketik. Jarang-jarang saya bisa beres course dalam satu hari.

Note

learning objective:

1. define AI
2. differentiate AI and augmented intelligence
3. describe 3 level of AI

AI = ability machine to learn pattern and make prediction augmented intelligence = it’s a sweet spot a AI can do and human can do it to like AI take large data, doing repetitive task, and do accurate human, generalize information, creativity and emotional intelligence

AI services is doing calculate not thinking, they a sophisticated calculating machine

machine learning = perform with acquire new data deep learning = calculation inspiration by neurons on human brain

AI, ML, DL On case sorting items

AI = fruit is categorized using an algorithm that scans the fruit and matches it with the correct label

ML = The fruit is sorted using an algorithm that identifies the types of fruit by various characteristics

DL = The fruit is sorted by an algorithm that recognizes different types of fruit from a variety of images

AI can do lot of thing now, simplest like correct you typing, recommendation food on your Go-Food apps, and specially your fyp on tiktok

Narrow AI (2010 - 2015) focus doing on single task like predicting next purchase of user Broad AI (Now) is a point between Narrow and General AI can doing larger task and versatile General AI ( 2050 - ) Can doing perform intellectual task like human ASI (artificial superintelligence) Machine self-aware

learning objective: describe history of AI, past to possible future

Era of Tabulation mean machines help humans to sort data into structures dark data = a huge chunk of data is hard to extract, is mess unsorted 2000 years ago Emperors Qin Shihuang used abacus to breakdown tax receipts and arrange them into category England at mid -1800s Charles Babbage and Ada Lovelace design “difference engine” to doing complex calculations using logarithms and trigonometry, but that never finised late 1880s Herman Hollerith, invented the recording if data on a machine-readable punched card

Era of programming is started at 1940s Scientist at University of Pennsylvania building Electrical Numerical Integrator and Computer (ENIAC), that could run more than one kind of instruction, now we called that program As example for program on that era, that helps astronauts problem on Apollo 13’s mission, this programable computer help astronauts safely back to Earth

Era of AI (Now) At 1956, a small group researcher, led by John McCarthy and Marvin Minsky launched revolution scientific research and coined term “artificial intelligence” AI Winter 1 Early 1970s, larger problem came because of 2 things, at that time calculating power is limited and also storage is limited to It took decades for technology catch up AI theory Late 1980s came another AI Winter AI Winter 2, that time over 300 AI companies shutdown or bankrupt In mid 1990s, forecast is changing to sunny. On 1997 IBM Deep Blue is winning at Chess match vs Garry Kasparov 2005, a Stanford University robot can drove itself on 131-mile desert trail 2011, IBM Watson defeated two grand champions in the game of Jeopardy

learning objectives:

1. differentiate between structured, semi-structured and unstructured data
2. Identify challenges that comes with working with unstructured data

Data = raw information Is maybe a fact, statistics, opinions or any kind 3 categorized of data: Structured data is categorized as quantitative data data can be structured into rows and columns Unstructured data or dark data typically categorized as qualitative data Semi-structured data is bridged between both data

learning objectives: 1. Define and describe machine learning 2. Describe how machine learning structures unstructured data 3. Describe how machine learning uses probabilistic calculation to solve problems

Two ways solve dark data problem 1. Programmable computer Need structured data upload to database. Data of all possible route, and add more data like weather and traffic condition. All of that is impossible to do, because that data would be revised every minutes 2. AI with machine learning Like climbing a tree, system would try a route. System will doing fork if route reaching branch until reached goal or dead-end. And choose shortest one. 2 advantages ML: ML can predict and ML can learning

Deterministic VS Probabilistic Deterministic is machine will flag things in term “YES” or “NO”, “ON” or “OFF”. Is binary thinking. To make answer true or false basically Probabilistic Machine cannot making answer true or false, it do probabilistic like this. “I confident this route is 84% is the shortest time”

learning objective: Describe three methods by which machine learning analyzes data

3 common methods on ML

1. Supervised learning
2. Unsupervised learning
3. Reinforcement learning

Supervised learning provide AI data to make accurate predictions. It need labeled data. Like a folder of dog pictured labeled as a dog data

Unsupervised learning Its like giving unlabeled data to algorithm, to make that data to be labeled.

Reinforcement learning It’s similar to supervised learning, but algorithm isn’t trained using sample data. This model learn by using trial and error. It will giving reward if right/positive. And give penalty if wrong/negative.

learning objective: Describe an ideal relationship between humans and machine learning

AI everywhere AI will move into all industries, from finance, to education, to healthcare. AI will increase productivity and enable new opportunities.

Deeper insights New technologies will sense, analyze, and understand things never before possible.

Engagement reimagined New forms of human-machine interaction and emerging technologies, such as conversational bots, will transform how humans engage with each other and with machines.

Personalization Machine interactions will be personalized for you, with new levels of detail and scale.

Instrumented planet Billions of sensors generating exabytes of data will open new possibilities for improving Earth’s safety, sustainability, and security.