If you're like me, you probably spend most of your time trying to avoid writing. It's not that I don't like writing, it's just that writing is hard.
Luckily for us, there's a new technology that can help make the process a lot easier.
🧵 Write Better, Faster With GPT-3
GPT-3 is a machine learning platform that can generate copy for you automatically. It's still in beta, but the results so far have been pretty impressive.
GPT-3 can be understood by understanding a few keywords.
1️⃣ Artificial Intelligence (AI)
2️⃣ Machine learning (ML)
3️⃣ Deep Learning (DL)
4️⃣ Language Model (LM)
5️⃣ Autoregressive (AR) Model
6️⃣ Natural Language Processing (NLP)
7️⃣ OpenAI
These terms will help you understand GPT-3.
1️⃣ Artificial Intelligence (AI)
AI is when machines do things that are normally done by people like learning and problem-solving.
2️⃣ Deep Learning (DL)
Deep learning is a type of machine learning where you teach a computer to learn by itself. This type of learning can be supervised, semi-supervised, or unsupervised.
3️⃣ Machine learning (ML)
Machine learning is a field of artificial intelligence that focuses on understanding and creating methods that "learn". This means that the methods get better at doing certain tasks as they get more data.
4️⃣ Language Model (LM)
A language model is a mathematical way of predicting how words will be said next, based on the probability of different word combinations.
Language models are probability distributions over a sequence of words.
They are used for many different tasks, like Part of Speech (PoS) Tagging, Machine Translation, Text Classification, Speech Recognition, Information Retrieval, and News Article Generation.
5️⃣ Autoregressive (AR) Model
An AR model is a way to describe a random process. This is used to help understand time-varying processes in things like nature and economics.
6️⃣ Natural Language Processing (NLP)
NLP is a way to help computers understand human language. It is a subfield of linguistics, computer science, artificial intelligence, and information engineering.
7️⃣ @OpenAI
OpenAI is a research lab that studies artificial intelligence. It was founded in 2015 and is funded by donations from people like @elonmusk and @Microsoft.
What is GPT-3 after all?
It stands for Generative Pre-trained Transformer 3 (GPT-3).
GPT-3 is a computer program that can create text that looks like it was written by a human. It's gaining popularity because it can also create code, stories, and poems.
GPT-3 has gained a lot of attraction in the area of natural language processing (sub-branch of data science).
Breakdown of Generative || Pre-trained || Transformer
Generative models are a type of statistical model which are used to create new data by understanding the relationships between different variables.
Pre-trained models are models that have already been trained on a large dataset. This allows them to be used for tasks where it would be difficult to train a model from scratch. A pre-trained model may not be 100% accurate, but it can save you time and improve performance.
The transformer model is a deep learning model that is used for tasks such as machine translation and text classification. It is designed to handle sequential data, such as text.
@openai a research lab in San Francisco, created a deep learning model that is 175 billion parameters (permutations and combinations) and can produce human-like text.
It was trained on large text datasets with hundreds of billions of words.
The language models prior to GPT-3 were designed to perform a specific NLP task, such as generating text, summarizing, or classifying. First, of its kind, GPT-3 is a generalized language model that can perform equally well on a wide range of NLP tasks.
How does GPT-3 work?
There are two kinds of machine learning: supervised and unsupervised.
Supervised learning is when you have a lot of data that is carefully labeled so the machine can learn how to produce outputs for particular inputs.
Unsupervised learning is when the machine is exposed to a lot of data but doesn't have any labels.
GPT-3 is an unsupervised learner and it learned how to write by analyzing a lot of unlabeled data, like Reddit posts, Wikipedia articles, and news articles.
If you are interested to try AI for writing, I recommend @heyjasperai and if you found this 🧵 useful, plz follow @weekly_concepts and RT the first tweet.
You can read the unrolled version of this thread here: typefully.com/weekly_concepts/RbToZhB