Tag: Artificial Intelligence

Uncategorized

WHAT DID VITALIK BUTERIN RECENTLY SAY ABOUT ARTIFICIAL INTELLIGENCE

Yogi Nelson

Namaste Yogis. Welcome to the Blockchain & AI Forum, where questions are answered, mostly correct! Here no question is too mundane. As a bonus, a proverb is also included. Today’s question comes from Art in San Francisco, CA. Art asks what does Vitalik Buterin think of artificial intelligence?

Art, you came to the right place. Before I explain Vitalik’s thoughts, let’s know Buterin first. Vitalik is a 29-year-old Russian Canadian computer programmer and a co-founder of Ethereum. Ethereum is a decentralized blockchain network, known for pioneering smart contracts, among other advancements in computation. From an early age Vitalik demonstrated an amazing ability to grasp mathematics and computer science. That’s not a surprise given his father was a computer scientist. Holy, like father, like son, Batman! Given his genius with computers, it makes sense you would ask what Vitalik makes of AI.

On November 23, Vitalik published his “My Techno Optimism” paper.
https://vitalik.eth.limo/general/2023/11/27/techno_optimism.html?ref=bankless.ghost.io Vitalik’s summary perspective is below followed by highlights from across the major points of his paper.

“I believe in a future that is vastly brighter than the present thanks to radically transformative technology, and I believe in humans and humanity. I reject the mentality that the best we should try to do is to keep the world roughly the same as today but with less greed and more public healthcare. However, I think that not just magnitude but also direction matters. There are certain types of technology that much more reliably make the world better than other types of technology. There are certain types of technology that could, if developed, mitigate the negative impacts of other types of technology. The world over-indexes on some directions of tech development, and under-indexes on others. We need active human intention to choose the directions that we want, as the formula of “maximize profit” will not arrive at them automatically.”

Vitalik Buterin, My Techno-Optisium

Buterin argues there are very high costs to delaying technological progress. He cites life expectancy improvements and the internet, as evidence for not holding back technological advancements. Vitalik makes the case for using advancements in AI to solve environmental challenges, including climate change.

Vitalik believes AI is fundamentally different from previous advancements in technology and therefore requires significantly more consideration. Vitalik says AI is the creation of a new and powerful brain whereas previous inventions were contraptions. “AI is a new type of mind that is rapidly gaining in intelligence, and it stands a serious chance of overtaking humans’ mental faculties and becoming the new apex species on the planet”, says Buterin. Vitalik cites scientists that are worried AI could wipe out humanity! Holy doomsday, Batman!

Buterin, is equally concerned with “digital authoritarianism”. Vitalik points out Open AI, with a mere 500 employees, services 100,000,000 customers with immensely powerful machines! Power could be centralized by just a handful of people one day, is Vitalik’s worry.

Vitalik pens a lengthy narrative regarding the e/acc movement. According to Vitalik, e/acc is “… fundamentally about an appreciation of the truly massive benefits of technological progress, and a desire to accelerate this trend to bring those benefits sooner.” Jeff Bezos is the most prominent advocate. In a point-by-point fashion, Buterin shares what he believes are the implications of e/acc across a range of topics, including: military technology, decentralized governance, cyber security, multi-planetary living, world government, to name a few topics.

We conclude today with this proverb from the Dominican Republic: With patience and calm, a donkey can climb a palm tree. Well said my Caribbean brethren!

Until next time,

Yogi Nelson

Uncategorized

STRONG ARTIFICIAL INTELLIGENCE, WEAK ARTIFICIAL INTELLIGENCE, WHATS THE DIFFERENCE?

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your blockchain and artificial intelligence technology questions are answered, mostly correct!   Here no question is too mundane.  As a bonus, a proverb is also included.  Today’s question, was sent by Sahir, from the Caribbean, and he wants to know what’s the difference between weak and strong artificial intelligence?

Sahir, you came to the right place.  At your age that’s a brilliant question!  Let’s start by explaining weak artificial intelligence (Weak AI).   Weak AI is also known as narrow AI.  Weak AI refers to AI systems that are designed and trained for a specific task or narrow sets of tasks. Weak AI systems excel in performing predefined tasks but are not remotely capable of true understanding or consciousness.  Weak AI is highly specialized and does not possess “human-like” cognitive abilities or general problem-solving capabilities beyond its narrow vision.  Think of Weak AI as a horse with blinders capable of only seeing what is directly nearby and in front and nothing else.  I’ll offer you examples. Weak AI includes virtual assistants, e.g., Siri or Alexa, recommendation algorithms used by streaming services such as Spotify or YouTube, and chatbots designed for specific customer service tasks, e.g., your financial institution customer help line.  Chatbots are mostly annoying and useless, do you agree? Time for Strong AI.

Strong AI, sometimes called Artificial General Intelligence (AGI), refers to AI systems with “human-level” intelligence and understanding.  Which human, I don’t know and afraid to ask!  Lol.  Strong AI systems can perform several intellectual tasks.  Strong AI systems can adapt to different domains and possess a “primitive” form of “consciousness or self-awareness” based on computations. Strong AI is a long-term goal of AI research and would require the development of AI systems that can reason, learn, understand, and adapt across a wide range of tasks and contexts.  Strong AI is currently a theoretical concept, and no AI system comes close to general intelligence.  Meaning you can relax–for now!  Lol. The best example of Strong AI, or perhaps somewhat Strong AI, is ChatGPT4 by Open AI. 

Let’s discuss a central feature of both Weak and Strong AI–prompts.  What is a prompt?  The term prompt is derived from Latin “promptus”, meaning ready for use.  In AI lexicon, a prompt is an instruction to the AI system.  A prompt could be simple, ‘Siri play disco music’, for instance, or complex, such as asking ChatGPT4 to explain Einstein’s Theory of Relativity. Holy prompt, Batman!  Sahir, can you identify which is the Weak AI and the Strong AI in these examples? 

Prompts are critical for two reasons.  First, prompts are how we communicate with AI systems.  AI systems can’t read our minds!  Only through, and thanks to, prompts, can humans interact with Weak and/or Strong AI systems.  In other words, prompts are portals into AI systems.  Second, the quality of AI interactions depends on precise prompts.  Unlike many human interactions, AI systems cannot “know what you mean”.  Therefore, AI systems depend on precise prompts.  My own experience with ChatGPT4 cements this understanding of AI systems.  Try it and give me your conclusion.

We stop here but not before acknowledging two sources for today’s post:  Microsoft’s Introduction to AI course and AI: The Future is Now by Nicholas Marty.  Today’s proverb is from Haiti, where they wisely say, “A leaking roof fools the sun but cannot fool the rain. 

Until Next Time,

Yogi Nelson

Uncategorized

WHAT IS THE FUTURE OF BLOCKCHAIN?

Yogi Nelson

Namaste Yogis. Welcome to the Blockchain & AI Forum, where your questions are answered, mostly correct! Here no question is too mundane. As a bonus, a proverb is also included. Today’s question comes from Ashley in Texas, who wants to know, what is the future of blockchain?

Ashley, you came to the right place. Blockchain technology has enormous potential; however, it must conquer immediate challenges. Let’s begin with five challenges facing blockchain technology today: 1) scalability; 2) initial set-up costs; 3) smooth transition; 4) consensus mechanisms; and 5) privacy and security. I’ll focus on a few solutions, commencing with “sharding”

Sharding is a process for handling large data sets. Sharding accelerates processing by dividing the computational workload and storage space issues. It ensures no single mode (computer) is responsible for processing the network’s transactional load. Sharding increases security through transparent processing on a decentralized network. Sharding is not alone; it has a friend known as sidechains; not to be confused with 2Chainz the rapper! Lol.

Sidechains are separate blockchains attached to the mainchain. Often, sidechains are deployed to test new software before joining the mainchain. Sidechains offer more security when moving digital assets from one blockchain to another and reduce the mainchain’s workload; hence, making the blockchain faster and more reliable. What else is under construction? Answer—state channels.

Despite their Orwellian name, state channels are not spooky. State channels basically “lock” the blockchain while participants agree amongst themselves off-chain. Once participants reach consensus the transaction is uploaded onto the mainchain for processing. Let’s keep rolling with “Roll-Ups”.

Rollups are scaling solutions. Roll-ups move computation off-chain while keeping transaction data on-chain. Keeping the data on-chain allows anyone to locally process all computation in the roll-up and detect fraud. There are two types of roll-ups: Optimistic and ZK. Optimistic roll-ups assume the data/transactions are correct. However, to ensure accuracy, transactions are not final for one week. During the data verification week anyone may submit fraud proof. If no honest individual spots an error, transactions are final. If errors are spotted the transactions can be reversed. What about ZK-Rollups? ZK-Rollups stands for zero knowledge. No, it does not mean the protocol is a know nothing, lol! It means every rollup batch contains a cryptographic hash, thus making transaction more secure. In other words, possession of the private key opens the file and nothing else is required. However, with the increase in security also comes computationally intensive processes. Time, money, and energy consumption.

Charles Hoskinson, Cardano founder, is fond of saying blockchains need their “wi-fi” moment. Said differently, blockchains must achieve interoperability. Imagine having to switch providers every moment to access wi-fi. What a mess! Blockchain interoperability equals users having a seamless integration of capabilities, communication with and between intermediaries, and greater decentralization. Get it done!

Let’s wrap it up with Ricardian contracts. A Ricardian contract is a mechanism to record a document as law and link it to other sectors, i.e., accounting. A Ricardian contract is responsible for executing contracts between two parties and recording the details in forms that are readable by humans and machines. Moreover, the dual abilities of Ricardian contracts equal superior user experiences when compared to smart contracts. Holy dual use!

Today’s “proverb” is curtesy of Yogi Berra to old time baseball star who once said, “it’s tough to make predictions, especially about the future”! lol. Nevertheless, I predict blockchain will evolve and eventually become mainstream technology with widespread applications. Thanks to the Emurgo Academy for providing the source material of today’s post. The art work is by Gerardo Cloquell and on display in Dorado, PR.

Until next time,

Yogi Nelson

Uncategorized

WHAT ARE THE 10 TERMS IN ARTIFICIAL INTELLIGENCE YOU SHOULD KNOW?

Yogi Nelson

Namaste Yogis. Welcome to the Blockchain & AI Forum, where your blockchain and artificial intelligence technology questions are answered, mostly correct! Here no question is too mundane. As a bonus, a proverb is also included. Today’s question was submitted by Alberto from Montebello, CA, and he wants to know the 10 terms in artificial intelligence (AI) that he should know.

Alberto, you came to the right place. Great question. Learning the vocabulary is an intelligent way to master the topic. Let’s roll it in alphabetical order.

Algorithm: I’m sure you have heard the term algorithm tossed about like a basketball. Let’s settle what it means; it’s not complicated. Algorithms are the step-by-step set of instructions followed by computers to perform a specific task. That’s it. Nothing more.

Artificial Intelligence (AI): is essentially teaching machines to learn from data and derive a variety of useful insights. AI is commonly defined as a simulation of human intelligence in machines that are programmed to think and learn like humans. Basically, AI enables computers to mimic human intelligence (hopefully a smart human) using logic.

Data Mining is all about discovery. In data mining patterns of knowledge from large data sets are discovered. Data discoveries are designed to assist with decision making, predictions, etc. Put on your computer hardhat and start digging!

Deep Learning (not to be confused with the Deep State, ha ha) is a specialized form of machine learning (as if machine learning is not sufficiently specialized!). Deep learning involves many layers of neural networks; hence, the label “deep”. (See below for neural networks.).

Machine Learning (ML): is a subset of AI that provide systems the ability to automatically learn and improve from the experience without being explicitly programmed. The key is not explicitly programmed.

Natural Language Processing (NLP): is the ability of a computer program to understand spoken and written human language. Holy linguist! However, human language is more than just sounds and we often have misunderstandings. Hence, it might take more time for software program to truly understand human language.

Neural Networks: a neural network is a computer system based on the human brain and nervous system. Which human I don’t know and afraid to ask! Ha, ha.

Reinforced learning is focused on finding a balance. A balance between what? A balance between exploration of new ground and exploitation of what is already known. Think of it as the yin/yang of machine learning.

Supervised Learning is a form of machine learning (see above). In supervised learning the system uses algorithm (see above) to identify patterns in the data and then “learn” from the patterns. Computer swamis call it supervised because the program is being directed by the instructions coded in the software. A good example of supervised learning is your email spam detector. Thank the universe for the spam detector, otherwise I may have fallen for a love letter from a 20-year-old Russian beauty queen! Ha, ha.

Unsupervised Learning is where the AI tells the boss to get loss! Ha, ha. Just kidding. In unsupervised learning the machine uses algorithms (see above) to analyze and cluster unlabeled data. In other words, in unsupervised learning the machine learns to uncover patterns without direct human direction. Transcribing audio is a good example.

A special acknowledgement to the Cardano Foundation Emurgo Academy, and these two books: AI Made Simple and Artificial Intelligence 2023.

Alberto, please remember those terms and the proverb from Suriname, where they say: You cannot fly with wings made of butter.

Until next time.

Yogi Nelson

Photo by Pixabay on Pexels.com
Uncategorized

ARE YOU SERIOUS, ARTIFICIAL INTELLIGENCE CAN HALLUCINATE?

Yogi Nelson

Namaste Yogis.   Welcome to the Blockchain & AI Forum, where your blockchain and artificial intelligence technology questions are answered, mostly correct!   Here no question is too mundane.  As a bonus, a proverb is also included.  Today’s question was submitted by Aaron from Dallas, and he wants to know if it’s true that artificial intelligence can hallucinate?

Aaron, you came to the right place.  The short answer is yes, artificial intelligence can and does “hallucinate” but don’t panic it doesn’t mean your AI has taken or was given LSD!  Ha, ha. Instead, let’s take a moment to define what it means for AI to hallucinate. 

Photo by Tara Winstead on Pexels.com

Within the context of AI, hallucinations are outputs, e.g., words, phrases, images, that are generated by the AI model (program) that are non-sensical or grammatically incorrect. In other words, the model made a mistake.  The error can be simple, i.e., a misspelled word.  Or, consequential.  For example, the hallucination could mean the model completely generated “facts”.  Put another way, it made up the facts.  Or, it may signify the model combined data that are individually correct but when fused together are wrong, non-sensical or grammatically incorrect.  Holy, computer glitch Batman!

Why would the AI model hallucinate, you ask?  I offer four reasons.  First, it could be the AI model is not trained on enough data.  Did you know, ChatGPT4 is running on mostly 2021 and 2022 data?  Second, perhaps the model was trained on noisy or dirty data.  The old garbage in and garbage out explanation!  Here is a third reason.  Maybe the model was given insufficient context.  Suppose you were only told the punch line of a joke without context.  The punch line within the context might be the most hilarious statement in Western civilization but would fall flat lacking context.  Last, an AI model may hallucinate if it is not provided sufficient parameters.  Imagine if you were told to guess a number between 1 – 10 compared to having to guess a number between 1 – 1000.  Would the difference in parameters affect your accuracy?  We both know the answer.

Now you know AI models may “hallucinate” and why.  With this new understanding of AI, remember to use common sense.  Remember AI isn’t perfect, subject to errors, and to bias.  Nevertheless, a great leap forward.  In keeping with today’s proverb from Jordan, I will end immediately with this proverb about people who won’t stop talking, they say, “He swallowed a radio”.

Until next time.

Yogi Nelson