dI
wrong prediction from Turing
"I Estimates of the storage capacity of the brain vary from 1010 to 1015 binary digits* ... I should be surprised if more than 109 was required for satisfactory playing of the imitation game." "I incline to the lower values and believe that only a very small fraction is used for the higher types of thinking. Most of it is probably used for the retention of visual impressions" - He predicted that by the year 2000, the answers of a computer would be indistinguishable from those of human beings, when asked questions by a human interrogator. This did not happen
The Agricultural Revolution
(10,000 BC): the transition of hunting and gathering to agriculture and settlement. People began to cultivate the land, plant crops, and raise animals.
The Industrial Revolution
(1760s-1840s It was the transition from home-run production to manufacturing on a larger scale in factories. This caused global trade, raise in income, and a higher standard of living.
The Information Revolution
(1960s-2010s) This is the period of change that positively affects the lives of humankind and computer technology is at the core of it. It includes all development of technologies that has led to the reduction of costs or ability to obtain, store, or process information.
Historical definitions of AI:
- "The science and engineering of making intelligent machines" (John McCarthy) - "The attempt to do things that would require intelligence if done by men" (Marvin Minsky)
technological revolution
- A set of interrelated breakthroughs that form a constellation of interdependent technologies (clusters of technology). - The two main feature of this are: 1) Strong interdependence of systems on innovation (Overlap of computers, software, phones, internet etc.) 2) They transform all economic aspects and society at large. - Since the industrial revolution, there have been 5 clusters of clusters: 1) Industrial revolution 2) Steam and rail 3) Steel and electricity 4) Oil, cars, and mass production 5) Information.
Jeff Bezos:
- Amazon CEO discusses the "gigantic" potential of artificial intelligence to change everything from shopping to self-driving cars.
Big data - The 4 V's:
- Big data consists of 4 attributes known as the 4V's: - Volume: - Velocity: - Variety: - Veracity:
3 games that were miles-stones in the history of AI:
- Computer automated games are often good testing ground for AI. - Each of these games used the same method and was merely an expansion of the one prior. - For each of these games, the computer computes in advance all possible outcomes of the game (Brute Force method) and therefore always has the optimal result. This computation of all possible results is known as the search space. The search space is often represented by an inverted tree. - Tic-Tac-Toe: Branch factor of 9. - Chess: Branch factor of 35. - Go: Branch factor of 250.
The computer as number processor, CPU, Pixel
- Computers understand information in the form of 0s and 1s as opposed to words. - The central processing unit (CPU) is the brain of the computer and sends control signals to other parts of the computer telling them what to do (like the brain of a human). - ASCII is a code that translates characters of a word in a decimal number which then translates into a number of 0s and 1s for the computer (A → 65 → 00100001). - In order to represent a picture, the computer uses pixels which are represented by numbers. For a color picture, there are 3 main colors used. They are green, red, and blue, each one having up to 255 pixels.
There are 4 basic skills needed to pass the Turing test:
- Natural Language Processing: The ability to communicate successfully in English. - Knowledge representation: The ability to interpret what information is fed to it. - Automated reasoning: The ability to use stored information to answer questions and draw conclusions. - Machine learning: The ability to adapt to new circumstances and detect patterns.
Bit VS Byte:
- The smallest form of digital data is a bit. - There are 8 bits in 1 byte. - Each byte is composed of 8 0s or 1s and each of those 0s or 1s is one bit. - 1 byte is the instructions a computer uses to represent a single letter.
7. (According to Thomas Kuhn: (scientific paradigm)
- The word paradigm originated with him, in the context of philosophy of science. In that context, a paradigm is the set of assumptions and beliefs, some even implicit, that are collectively held at a certain period. -
How Facebook optimizes data:
- This issue of needed to filter out data is an example of the optimization problem. - by trying to optimize the amount of time each user spends online, escentially trapping them, which increases revenues. - developed sophisticated algorithms to solve this challenging optimization problem, taking into account different factors including the estimated interest of the user in each post, the post attractiveness for other users, information about the creator of the post, the type of post (picture, text) and its recency. These sophisticated algorithms adapt to the constant environment, and applies machine-learning tools
A/B Testing
-goals is to identify possible changes to web pages that increase a measure of interest (a KPI). - It is done through a controlled experiment where there is an independant and dependant variable. For example, to test the UX of an application, a company may change the color of a button (Independent variable) to see which users like better and therefore, click more (dependant variable), leading to an increase in revenues (KPI).
Database and Data cloud:
-is a collection of information that is organized so that it can easily be accessed, managed, and updated. - Relational databases are made up of a set of tables with data that fits into a predefined category. The Structured Query Language (SQL) is the standard of relational databases. An example of a relational database is shown below. - Today, most data is stored in the cloud. The cloud refers to software and services that run on the internet (Google Drive or Dropbox). - The cloud exists in physical form, in a 'data center', in some rural and remote areas.
Machine Language (Machine Code)
0 and 1
Third feature: combinatorial innovation
1) Steady exponential improvement (computing power, different technologies) 2) Size and breadth of digital information 3) Re-combinational digital Innovation
. 10/10 and 1/1 year rules according to Johnsone
10/10 rule: It usually takes 10 years to develop an innovative platforms and 10 years for mass adoption. - Today, due to the internet and fast adoption rates the 10/10 rules has become the 1/1 rule. It only takes only 1 year to develop an innovative platform and 1 year for mass adoption. YouTube is a prime example of this.
Types of companies :
1: Assets Builders 2 Service Providers 3 Technology creators 4 Network orchestrators
3 features of big data
3 features of big data are: - Big: N= all. In the book Freakonomics, they were able to detect a cheating pattern in sumo wrestlers. Sumo tournaments have 15 rounds and when the score is 7/7 there is an incentive to cheat. The time between rounds allows for the less capable player to bribe the more capable player to let him win. This trend was only found because they analysing the N= all. - Messy: The bigger the data being analyzed, the less exact it gets and there is room for some error or inconsistencies. - - Example: if you're looking at a post on Facebook from a friend, you will see if it has 5 likes or 123 likes. When looking at a post from someone famous however, you will see a broad range of 300K likes or 1.5M likes. In the 1950s, there was research done to translate large amounts of text from Russian to English. Through Translation Via statistical measure, IBM applied a machine learning algorithm to learn the connection between phrases and sentences in the two languages, that is, translating with context. However, after an initial success, project abandoned, as it did not yield the expected results. Google translator used this same approach and created the best algorithmic translation so far using 2 differences. They used a larger amount of messy data which solved a problem that smaller and cleaner data could not. - Good enough: We can gain insight from correlation and not causation. It is useful to discover patterns in data without discovering what causes them. - We can use correlation to find patterns. For example, the Manhattan electrical grid is a network of over 21,000 miles of underground cable. By having records of past explosions, a model has been created to predict future explosions.
Three human skills that are not well-defined, and are hard of AI
: - Visual perception: The brain uses additional information that isn't given when extracting meaning from an image. For example, the human eye uses depth perception while a computer does not. - Walking: It took ten years to develop prototypes that resembled artificial walking. Only in recent years have robots been able to achieve adaptation of walking in natural environments (being able to walk when encountering snow for example). - Language: Computers have a hard time understanding context of words. - Example: The sentence "Time flies like an arrow" would usually be understood to humans as time moving quickly. To a computer however, time may understand time to be a type of fly that enjoys arrows.
Classification
= supervised learning with discrete outputs example - fraud detection: OK or FRAUD
Artificial Neural Networks (ANN)
A common tool in machine learning are loosely based on biological neural networks. are computational structures that draw inspiration from biological neural network. A single artificial neuron is a function that get several inputs, perform a computation, and transmit an output. This changes around 2012, with the development of a specific set of "this" called deep neural networks. Deep neural networks are currently a very hot topics in computer science and other fields. In particular, this technology seems to be able to solve problems that were not solved before, for example problems of object recognition in computer vision.
learning
A process for gaining knowledge or improved performance based on EXPERIENCE.
- Weak (narrow) AI:
A system that focuses on only one task (EX: speech recognition).
Artificial general intelligence:
A system that has an artificial intelligence that can successfully perform any intellectual task a human being can.
Technological economic paradigm (TEP):
After a TR, the paradigm that sets in is called the new Technological Economic Paradigm (TEP). - The new TEP sets a new common ground, a new set of assumptions and expectations. - These can include commonly held views on the expected change of costs (Moore's Law), the possible space of new innovation (50'-60: "cars getting BIGGER and faster"; Today: "star-up companies are the best!") and new organizational principles (Taylor's assembly-line method which spread after the Ford T-model). - TEP only reaches the public sector (government) after 20-30 years. - This may be cause high pressure to adapt and due to organizational inertia, which is the tendency of a mature organization to continue on its current trajectory.
Order of bytes from least to greatest
Byte Kilobyte Megabyte Gigabyte Terabyte
There are 4 basic skills needed to pass the Turing test: There are 2 more skills needed to pass the total (embodied) test: -
Computer vision: The ability to perceive images and make sense of them. - Robotics: The ability to manipulate objects and move around.
The definition of data and digital data
Data is a set of values of qualitative or quantitative variables. - The difference between data and digital data is how it is stored. - Data: tax lists, weapon inventories, food supplies - Digital data: captured in a digital form on a computer, and can be replicated easily - It is also important to distinguish between digital data and information. For example, a picture of a textbook is digital data. It is stored as a picture and therefore, does not have any actual meaning. To convert it into information, we can perform optical character recognition (OCR) and extract the words from the picture and turn it into meaningful information. - In this course data, digital data, and information are often synonymous.
First feature: Exponential Growth
First Machine Age: mainly about new source of power to change something in the physical world (move some atoms around) Steam, Coal, Oil ● We have entered a new era, were new technologies provide new ways to control, to operate power. At the root of this era are two related inventions - the Computer and Software. The current technological revolution is about non-physical inventions: ideas and information. ● "Computers and other digital advances are doing for mental power— the ability to use our brains to understand and shape our environments— what the steam engine and its descendants did for muscle power.
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: Mindset
From closed to open Ex. The restructuring of google resulted in the Alphabet Inc. Over time companies get comfortable with only making incremental changes But in the technology industry, you must be a bit uncomfortable to stay relevant *Innovative science demands leaps into the unknown
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: Revenues
From transactions to subscriptions Types of Subscriptions 1. Things (ex. Boxes) 2. Services (ex maintenance) 3. Information (ex adobe) 4. Networks (ex LinkedIn)
Technology creators
Ideas - make one , sell many - , delivers value through ideas . Develops and sell intellectual property , software , analytics , pharmaceuticals & biotechnology ( ex : Microsoft , Pfizer )
features
In both supervised and unsupervised learning, the computer is given a set of inputs For example, for classifications of 'apples' and 'oranges' a good set of"this" can be 'color' and 'weight'. A machine that measures the colors and the weight of a fruit will be able to learn that apples are usually lighter, and have tend to be green to red, and use these features to differentiate between apples and oranges. When a feature corresponds to a numeric value (such as weight) we can think of it as a dimension in the mathematical sense. The space that is defined by the different dimensions is the search space of the problem. In class, we discussed a two-dimensional machine learning problem.
The importance of innovation by Joseph Schumpeter's (1912)
Innovation is the main force of economic growth Schumpeter argued that capitalism can only be understood as an evolutionary process of continuous innovation and creative destruction.
Jobs that will cease to exist for humans due to atomization and robotics
Machines can do everything humans can do so people are losing jobs. - There are many new jobs coming out as well such as application developers, video bloggers, drone operators etc. - 47% of total employment in the US is at risk in the next decade or two. - Many white-collar jobs will be fully automated soon (Tellers, brokerage clerks, insurance clerks, packaging). - Blue color jobs are in lower risk than white color jobs (shampooers, barbers, floor layers, systems assemblers, repair workers). Making machines that perform physical tasks that in general don't bring in high income is more demanding and less rewarding for companies. Jobs with low risk of automation are ones that are complex, creative, physical, or empathy related such as nutritionists, vacation managers, choreographers, physicians, psychologists, dentists etc.
the advantages of the societies that had the agriculture revolution
More food => more people Labor division Agriculture Public works Warriors => Conquest Land, protect resources Technology development Metallurgy Tools, weapons Writing
Effects of technological revolutions of employment
Negative impacts of first industrial age: It take society time to realize the negative impacts of a technological revolution. At some point public opinion, regulations, kicks in. After industrial revolution: child labor >> time gap >> rules against Nowadays: 'flat world' ("Facebook and Google are great!") >> time gap >> "OMG! Privacy, Fake news.." => GDPR ettc SOOT in London : challenges can be met (child labor abolished; env. Regulation set), need to understand the ones for our generation
Netflix and Blockbuster example
Netflix started as an alternative to regular DVD rental shops such as Blockbuster. The innovation of streaming (watching movies over the interne), and Netflix innovative approach, such as using big data to personalize recommendation (will discuss later in the course), led eventually to the destruction of the DVD rental business. Blockbuster, which rejected an offer to buy Netflix for $50M in 2000, went bankrupt in 2010.
incremental innovation
Once a paradigm sets in, it dictates a direction for more technological improvements, leading to a sequence of incremental innovation. These are smaller, less revolutionary innovations, which follow a more expected trajectory. If the first iPhone design was a radical innovation, the move from, say, iPhone 5 to 6 is an incremental innova ● An example of mix-and-remix is the invention of the iPhone: With the invention of the iPhone we can see an example of how Steve Jobs adapted to the accelerating changes in order to succeed. Steve Jobs took technologies that existed and combined them to one simple and user-friendly interface. The original iPhone was a described as a combination of three technologies: a phone, an internet browser and an iPod.
Service providers
People - hire one , sell one - delivers value through skilled people . Workers who provide services to customers ( JP Morgan , Accenture )
Network orchestrators
Relationships - many make , many sell- low marginal cost . Delivers value through connectivity . Create platforms where people can interact with one another ( including selling products , building relationships , sharing advice , giving reviews celebrating and more (ex : Ebay , Visa ) . It provides access to different types of capitals through relationships (ideas , things , people , more relationships ) In the lead in the economic measures
The innovation of the first iphone
Smartphone with integrated revolutionary mobile ,widescreen Ipod with touch controls, Breaktrought Internet communication device - Steve Jobs introduced the first iPhone in 2007. - He removed physical keyboards which just take up space and created a multitouch interface. - He included a web browser and the ability to sync with iTunes
examples of negative impact on industrial
Society took too long to realize the regulations needed such rules against child labor. He compares this to the way that right now technology is interfering with privacy/ fake news is everywhere but how we will only realize it in retrospect. - Jobs for farmers have been diminished as machines have taken their place. Example in the beginning of the 19thcentury, 86% of the workers in the US were employed in agriculture. Today (beginning of 21stcentury), only 2% percent of the people are doing so
3 levels of AI:
Strong AI: - Artificial general intelligence: - Weak (narrow) AI: Most AI are weak AI and don't attempt to achieve strong AI.
The five fingers rule (components of frame setting)
Targets/goals Topics Times Rules Roles
TEN PRINCIPLES FOR NETWORK ORCHESTRATION (only 8 of them )
Technology ASSETS STRATEGY Leadership From customers to contributors revenues from employe to partner mindset
3 main technological revolutions
The Agricultural Revolution (10,000 BC) The Industrial Revolution (1760s-1840s The Information Revolution(1960s-2010s)
4V's:Variety:
The data comes from different sources, both structured and unstructured. Structured data is data that is well defined such as money which will always have numbers that contain 2 decimals. Unstructured data has no rules and can vary from being a text, audio, pictures etc.
the main theory of guns steel and germs
The effect of geography • Best plants for domestication (5,000 years earlier than Americas) • Best animals for domestication • East-West orientation Climate similar => Crops spread easily
. the main question of guns steel and germs
The opening question of the book is "Why did wealth and power become distributed as they now are rather than in some other way?" - The author, Jared Diamond, attempts to explain the current state of the world by looking for the roots of the first big technological revolution. This is the change of humanity from hunters/ gatherers to farmers. - As a starting point for his analysis, he focuses on one historic event - how the Spaniard Pizzaro took over the entire Inca empire (1533) with few hundred soldiers and horses? What gave them the ability to do so? - The short answer is that the Spaniard had Guns, Germs (that actually killed most of the Native American, through different diseases that were not known in the Americas) and Steel due to advances in technology. - Diamond's provocative argument is that the landmass of Eurasia had geographical (best plants and animals for domestication) and biological advantages that allowed the Agricultural Revolution to happen in that area of the world first.
4V's: - Volume:
The size of the data should range from Terabytes to Exabytes. Any smaller does not constitute as a large enough volume. In big data, N= all (usually messy and diverse). In a sample, N= small (clean and uniform).
from Force to Control - (the second Machine Age )
The steam engine started it all. More than anything else, it allowed us to overcome the limitations of muscle power, human and animal, and generate massive amounts of useful energy at will. This led to factories and mass production, to railways and mass transportation. It led, in other words, to modern life. It involves the automation of many cognitive tasks that make humans and software-driven machines substitutes, rather than complements. - It is contrasted with the "First Machine Age", or Industrial Revolution, which helped make labor and machines complementary.
Why some inventions are not innovations
The study of innovation: technology + economy + society PARADIGM
Assets Builders
Things - make one , sell one - high marginal cost , delivers value through the use of physical things ( ex: walmart , airbus )
reinforcement learning
This method focuses on lies between unsupervised and supervised learning. an agent (a robot, or computer simulation) interacts with its environment, and is given a reward function that it tries to optimize. For example, the agent can try to walk (using different parameters for its algorithm), gets positive feedback when manage to walk a few steps, and keep using these parameters, and then try to improve more in an iterative manner. This method is based on understanding on how people and animals learn new behaviors.
4V's:Velocity:
The frequency of incoming data that needs to be processed such as in the Facebook feed example above.
-4V's: Veracity:
Uncertainty due to data inconsistency and incompleteness, ambiguities, latency, deception etc.
Proof of rapid technological changes
We live in a period of accelerated change and a speedy growth rate. For example, the spindle (weaving tool) took 120 years to spread while the internet and iPhone only took 10 years. Additionally, it took 62 years for 50 million people to use cars but only 3 years for 50 million people to use Facebook.
Machine learning
a field of computer science that gives computers the ability to learn without being explicitly programmed.
Algorithms
a set of well defined instructions. - In order to program this, we write computer code which is a set of formal instructions that can be understood by computers. - High-level programming languages use English words to make it more understandable to the programmer. - A translator then interprets high-level code into machine language (0s and 1s) so the computer can understand it.
supervised machine learning
a system is trained with data that has been labelled. The labels categorize each data point into one or more groups, such as 'apples' or 'oranges'. The system learns how this data - known as training data - is structured, and uses this to predict the categories of new - or 'test' - data.
Network centric companies
almost no physical assets —> are becoming more popular Example : Uber vs Herz AirBnB vs Starwood WeChat vs Alibaba vs Walmart focuses on co-creation provides digital platforms applied shared assets gains from big data insights
Target predicts teen pregnancy
an often-told story is about Target (a general store in the US), that used machine learning to predict their buyers' future purchases. The company correctly predicted that a teen-age daughter is pregnant (based, for example, on new items bought such as Folic Acid), to the annoyance of her father: This story might be an urban legend, but the fact that it is told and re-told suggest that it touches upon something that is happening now: big companies collect a lot of data on us, and they use tools, such as machine learning algorithm, to know private things on us, things that sometime we do not know.
Digital Platforms :
are at of orchestrators A product is useless without a platform —> Facebook succeeded because they had other people build a constellation of products —>Google + thought differently and failed , they tried to predict what people want and deliver it t them- should've thought like fb & had others seen what they each want - fb is different fr everyone , whereas Google + is the same for everyone
2 milestones in AI
are in 1986 with artificial neural networks and in 2012 when ImageNet had a breakthrough in deep learning with their visual recognition software.
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: From employees to partners
because of the differences in relationships in the 21st century, there are changes to customers and employees. More options and more information=easier to switch (supplier; job) 34% if US workforce freelance in some way
Type of machine learning problems
classification regression clustering
How Google sets your feed? Filter Bubble
collects data on individuals based on their clicks, search history, profiles etc and uses it to alter the search results per individual. This means that if 2 people use the exact same thing, they will be fed results that fit what Google thinks they want to see. - This prevents expansion of opinions and research.
Digital Networks
expand / less effort —> Facebook - 6 years - over 500 million users New members = new potential friends Ebay —> each seller is a potential customer
Turing Test
for AI ('imitation game') Developed by Alan Turing in 1950, is a test of a machine's ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human. .
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: Leadership
from commander to gardener Relationship change over time between employees and customers 21st Century Relationships • More options (ex. To buy; switch jobs) • More information • More participation (ex. Co-creation; brand review) Leaders must create broad environments instead of command and control
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: Strategy
from operator to re- allocator Companies allocate And some re allocate ( innovative) : lower year by year correlation Ex : IBM - let go of laptops - Lenova Some don't (conservative) -perfect year by year correlation: ex: Kodak not letting go of film industry. Higher levels of reallocation= higher returns
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: ASSETS
from tangible to intangible Categories : 1) information ( eg relevant people ) 2) relationship and access 3) trust Now management practices are needed to handle intangible assets ( handling customer satisfaction) - Customers yield greatest power -Facebook has positive customer satisfaction comcast has negative customer satisfaction - won't just accept a customer leaving the company , won't stop asking why CRM= customer relationship management
hokey stick figure
graph is one that shows a relatively flat period (the shaft) followed by a sharp increase (the blade). ● concept originated in debate on climate change, and is now a popular term used to describe a phenomenon that start to grow exponentially, or in other words, to 'explode'. It is widely used in start-up sale pitches, where founder hope to convince investors that their business idea will take over the market, with 'exploding' market shares and revenues
Main idea of TED talk of Hans Rosling
he says that the greatest invention of the industrial revolution is the washing machine due to the availability of time it created.He makes the case for the washing machine. With newly designed graphics from Gapminder, he shows us the magic that pops up when economic growth and electricity turn a boring wash day into an intellectual day of reading.
Machine learning
is a class of methods that attempt to enable computers to be adaptive and learn from examples. these algorithms are the heart of many successful AI applications such as Natural Language Processing (NLP).
Biological Neural Networks
is a collection of thousands and millions of interconnected neurons. Understanding biological neural networks is at the core of the current paradigm of neuroscience, the science that study the nervous system, that is, the brain, mind and behavior of human and other animals.
uncanny valley
is a common unsettling feeling people experience when androids (humanoid robots) and audio/visual simulations closely resemble humans in many respects but are not quite convincingly realistic
Key performance indicator:
is a measurable value that demonstrates how effectively a company is achieving key business objectives. Some possible of this are sales, user experience, customer satisfaction, increase of subscribers etc.
Natural Language Processing (NLP
is a type of machine learning algorithm which is inspired by the main computational structure of the human brain, the biological neural networks.
Google Duplex
is an AI calling system for accomplishing real world tasks over the phone. - Example: The following is a recorded phone conversion of "this" making a reservation at a restaurant
Alan Turing
is often called the father of modern computing. He was a brilliant mathematician and logician. - He developed the idea of the modern computer and artificial intelligence. - During the Second World War he worked for the government breaking enemy's codes and Churchill said he shortened the war by two years.
deep learning
is simply systems that perform machine learning using deep neural networks. Many of the current advances in different field of AI is based on deep learning.
Farecast using big data:
is the first company that used big data. - Oren Etzioni created a predictive software that was used to determine future prices of airfare by collecting historical data. - they uses machine learning to advise whether a person should buy a given ticket immediately or wait to get a better deal.
Product manager UX
is the intersection of business (strategy), technology (execution), and User experience (user understanding) of a product and helps companies focus on the big picture. - is responsible for delivering high quality digital products. - helps the company decide what features or products should be made, ensure that it happens, and reports back feedback on user response.
The AI Effect:
is the phenomenon in which once a problem is solved in AI, it is no longer True AI but merely AI engineering.
Artificial intelligence (AI)
is the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
Unsupervised learning
learning is learning without labels. It aims to detect the characteristics that make data points more or less similar to each other, for example by creating clusters and assigning data to these clusters.
Taxes on Digital Technologies -
mobile , social media , big data + analytics , interact of things (I.T)
Connections in a network
n(n-1)/2 2 phones : 1 connection 5 phones : 10 connections
Lean Startup
one that looks to reduce costs and has rapid testing/ short development cycles (opposite of government sectors).
Deep learning:
part of a broader family of machine learning methods based on artificial neural networks (enabling computers to create predictions on complex data).
TEN PRINCIPLES FOR NETWORK ORCHESTRATION Technology
physical to digital - —> brick & mortar stores vs E-commerce Taxes - Digital Technologies - Digital Platforms :
Conversion rate:
regarding a website is the amount of visitors to that website that compete the desired goal (usually a purchase) out of the total amount of visitors to the website. optimization is the approach of how to maximize the conversion rate. In order to do so, usually analysis of user interactions takes place (how they move through the site, what's stopping them from completing the goal etc.).
The uncanny valley:
s a common unsettling feeling people experience when androids (humanoid robots) and audio/visual simulations closely resemble humans in many respects but are not quite convincingly realistic.
1. Wrong predictions by Ray Kurzweill
says that machine intelligence will be infinitely more powerful than all human intelligence combined. - Afterwards, he predicts that intelligence will radiate outward from the planet until it saturates the universe. - Previous wrong predictions from 1999 book include: 1) That machines with human-level intelligence will be available from affordable computing devices within a couple of decades. 2) Nanotechnology will augment our bodies and cure cancer. 3) Humans will be able to connect to computers via direct neural interfaces or live full time in virtual reality.
frame setting
sets a common frame for all participants - what are the goals, time line etc, and create a clear structure for collaboration.
radical innovation
starts a new change. It is a big change, a jump from the current status-quo. It is being adopted slowly at first (by early adopters), and different designs are competing with each other. Ad some point, a dominant designsettles in, and the innovation starts to take over the market. An example of this process was the design of the first personal computers in the 80'. While at the beginning of the decade (when the personal computer was new) there a lot of competing design for hardware and software (Atari, Commodore, Amiga, Apple II,..), by the late 80' the majority of users have adopted the IBM PC model (with Intel microprocessors and Microsoft operation-system)
Law of increasing Returns
success is self reinforcing in economic network : the value of the network is in the connections (edges ) each new mode adds to the number of connections exponentially
Regression
supervised learning with continuous outputs example - house prices in this class
Singularity
the point in time where machine intelligence will be infinitely more powerful than all human intelligence combined. also the point at which machines intelligence and humans would merge, changing humanity as we know it.
creative destruction
the process of industrial mutation that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one. This is a concept made by Schumpeter An example of this is Netflix VS Blockbuster. As Netflix came out with streaming, the DVD rental business was destroyed, and Blockbuster went bankrupt.
examples of technological paradigm
the shift of size of microprocessors (computers) to be smaller and smaller. But in the 50s/60s, they were assumed to get bigger and bigger until there was a paradigm shift. It is kind of like the built-in standard/assumption we work off of. - Moore's Law is an example of this.
Neuron
type biological cell. which compose the nervous system that include the brain, the spinal cord, and the peripheral nervous system.
• Clustering
unsupervised learning to find similar sub-groups in the data
Amazon Go:
used AI and deep learning to create their store.
Moore's Law (Gordon Moore)
was one of the founders of Intel. is the principle that the speed and capability of computers can be expected to double every year, as a result of increases in the number of transistors a microchip can contain. is both consistent and broad; it's been in force for a long time and applies to many types of digital processes.. applies to many digital processes (eg, hard-disk cost efficiency, measured in Gigabytes / Dollar), over many years.
Research Cloud:
when experiments continue to fail, you reach a cloud- you get lost in the cloud, but sometimes to get to a different conclusion than you had planned- must remember how
Strong AI:
would be a system with a mind in the same sense that a human has a mind.
Possible sources of innovation + examples
· new products · new production methods · new markets new forms of organization new products, new production methods (Ford T assembly line), new markets, and new forms of organization (Wikipedia). These all result in new demand which creates wealth.
Physical networks :
—> highway systems - 35 years - 425$ billions As it spread , physical growth - oil , machines , food , people , services , ideas
TEN PRINCIPLES FOR NETWORK ORCHESTRATION: From Customers to Contributors
• Talent: external designer community-don't need to find, hire, or train • Insights: ex. All printed t shirts will be popular • Agility: systems adapt to changes in styles • Scalability: ex. External talents; online voting; easier to have more/less designs • Access: co creators who act as promoters in other networks • Affinity: co creation leads to deep relationships with designers and voters *Example: Legos can give an idea for a lego set. They hired an outsider as their CEO-this is what the company needed to get back on their feet- they fixed the business before they "rebuilt" it
the main advantaged that the Spanish had upon the inka tribe
• steel armor • swords • horse mounted cavalry • guns • .. And: deadly diseases that the local were not immune to
. Innovative product according to Ashton Kucher
• the idea is audacious • the idea need to have "an addressable, total gigantic market" • the idea sounds at first almost non-sensible, for example, in AirBnB, the idea that people will sleep on strangers' living room sounds idiotic
. Graphical User Interface (GUI), History of GUI
•- This is the part that we see (computer screen, phone screen) as opposed to where the calculations are actually taking place. - We only see the part that is pleasing to the eye. - The first GUI was Engelberg 1968. - In 1981 Xerox Star was produced. - In 1982 Apple Macintosh was produced. - Mass adoption of this was in 1990 with Windows 3. ADD = DEF OF G.U.I, H.C.I The 90-minute presentation essentially demonstrated almost all the fundamental elements of modern personal computing: windows, hypertext, graphics, efficient navigation and command input, video conferencing, the computer mouse, word processing, dynamic file linking, revision control, and a collaborative real-time editor (collaborative work).
the tail of the invention of chess
● According to the legend, a mathematician invented chess for a king in India who loved gamed. Asked for what he wants in return the mathematician said: I want one grain of rice on the first square, than double this (2) on the second one, than double this (4) in the third one and so forth until the last 64thsquare. The kind was happy to grant this easy wish. However, as his servants advanced in the board, more and more rice was needed. Around half of the chess board, the amount 'exploded', and all the grain of rice in the kingdom were not enough. "You have provided me with such a great game and yet I cannot fulfill your simple wish. You are indeed a genius." said the King and offered to make the mathematician his top most advisor instead
Why digital data is unique
● Exact replicate ● (almost) zero cost for replicate ● (almost) zero cost and time to distribute => PERFECT + FREE + INSTANT
the difference between the internet and www
● Internet is NOT equal the Web (or formally, the World-Wide-Web). ● The internet is the infrastructure - all the computers that are connected together. The Web is one type (albeit major) of information sharing model - the one we are mainly exposed to through web browsing. The internet also supports other information sharing model - for example, email or FTP.
the difference between innovation and invention
● Invention = new discovery in science and technology ● Innovation = commercial introduction of new product, or "new combination" Not all technological inventions enter the market due to economical and social reasons Need to be: (1) Economically profitable (2) Socially accepted - Innovation is the introduction of a new product or new combination while invention is a new discovery in science and technology.
Linear and Exponential functions
● Star-trek Tribbles: Multiply every DAY Liner growth can be higher at the beginning, but exponential growth takes over in the long run
ambient technology and examples
● a technology that is so intuitive and effortless to use, that it 'fades into the background', and people use it implicitly. Such technology is termed ambient technology, in a similar way the ambient music is music that is in the background, it affects us but we hardly notice it. ● The term comes from the definition of Ubiquitous Computing= computing anytime and everywhere ● EX : ipad ipod
the second feature of the modern age - combined innovation
● all technology is build in layers. In order to become a farmer you need to domesticate cows, for this you need ropes, buckets, the ability to create sheds - each part is a technology that you need to master.
innovation as mix and remix
● much of our current technology is based on ideas and information, more and more new innovation are developed by mix and re-mix: combining existing elements in new ways. ● A canonical example is the development of WAZE. ○ Waze can be considered as a recombination of: GPS system + Data transmission device (mobile) + Social network
