Discover and read the best of Twitter Threads about #gpt3
Most recents (24)
🏁 Here we go with Part 3
While I was writing Part 2 I finally got my #gpt4 API access 🥳
Using the API #chatgpt, it answers at least 3x faster, cool
So we're back on track from the @openapi playground while literally siphoning my 💵
Some thoughts about that...
1/11
While I was writing Part 2 I finally got my #gpt4 API access 🥳
Using the API #chatgpt, it answers at least 3x faster, cool
So we're back on track from the @openapi playground while literally siphoning my 💵
Some thoughts about that...
1/11
#gpt4 is ~20x more expensive than #gpt3
After 10 minutes I ended up maxing the context (what's taken into account for the response) and I was already at $2.18
May seem small but it really depends on your situation 🌎
2/11
After 10 minutes I ended up maxing the context (what's taken into account for the response) and I was already at $2.18
May seem small but it really depends on your situation 🌎
2/11
When you figure out the Purchasing Power Parity between countries, or even within the same country, those $2 for 20 minutes of #gpt4 can represent a lot and is out of reach for some
Hopefully there are 4 things that'll likely happen 👇
3/11
Hopefully there are 4 things that'll likely happen 👇
3/11
I already love #GPT4
I love it so much that I'm getting it to turn itself into a fully functional pair programming assistant inside of a #vscode extension
4 hours in, it’s already writing 90% of its own code
To do this, I'm teaching it a new language
1/? 👇
I love it so much that I'm getting it to turn itself into a fully functional pair programming assistant inside of a #vscode extension
4 hours in, it’s already writing 90% of its own code
To do this, I'm teaching it a new language
1/? 👇
#GPT4 is surely amazing, but its top current limitation is interactivity. Out of the box, #chatgpt4 has no way of directly with outside systems, which is a drag
The thing is, you can actually teach it to
But first, let’s analyze the way we interact with the chat website
2/?
The thing is, you can actually teach it to
But first, let’s analyze the way we interact with the chat website
2/?
When asking for code or shell commands, #GPT4 always responds with a code block
There are a couple of ways I interact with those
1. For commands, select, copy, paste. Line by line
2. For files, I click “copy” button
3. For partial files: select, copy, paste
3/?
There are a couple of ways I interact with those
1. For commands, select, copy, paste. Line by line
2. For files, I click “copy” button
3. For partial files: select, copy, paste
3/?
all my friends who read a lot when young are successful now
i think reading is probably the more important thing you can do in elementary school
if ur brain is like an LLM, you should maximize the # of tokens it sees while training
i think reading is probably the more important thing you can do in elementary school
if ur brain is like an LLM, you should maximize the # of tokens it sees while training
i read so much in elementary school that writing is very easy for me. i liken it to next token prediction.
SAT/ACT grammar came very easy for me, i just went off of what sounded right. idioms as well, even though my family isn't native english speaker. but id read all them
SAT/ACT grammar came very easy for me, i just went off of what sounded right. idioms as well, even though my family isn't native english speaker. but id read all them
⚡️ #GPT4 est là depuis moins de 24 h et les applications captivantes se multiplient !
Je les liste dans ce thread au fur et à mesure de mes découvertes !
⬇️⬇️
Je les liste dans ce thread au fur et à mesure de mes découvertes !
⬇️⬇️
🔥 Les essais percutants de @Yeca, orienté #SEO bien sûr ! #SEOEnAction
Que vous pouvez retrouver dans ce thread :
Que vous pouvez retrouver dans ce thread :
一些关于#ChatGPT4的信息：
和前文一样，本文也将使用ChatGPT工具链来解读相关消息。
首先我们通过New bing（如果你还没有可以切换成日本IP申请，大概率秒过），来询问它有关ChatGPT的消息：
可以看出，主要要点就是下周发布，可以生成动画，但我希望有更多的信息。
和前文一样，本文也将使用ChatGPT工具链来解读相关消息。
首先我们通过New bing（如果你还没有可以切换成日本IP申请，大概率秒过），来询问它有关ChatGPT的消息：
可以看出，主要要点就是下周发布，可以生成动画，但我希望有更多的信息。
于是我选了他给出的其中一个参考网站，得到进一步信息（此处用的是 @yetone 的翻译插件，前文中已提到之后不会再提），下面继续解读这几条
ICYMI - I made an AI-generated deck builder over the weekend 🔥
3 versions:
📊 Startup Description > Pitch Deck
👩🏫 Prompt > Lesson Slides (#education)
📚 Topic > Illustrated children's story
Check them out 🔻
3 versions:
📊 Startup Description > Pitch Deck
👩🏫 Prompt > Lesson Slides (#education)
📚 Topic > Illustrated children's story
Check them out 🔻
📊 Startup Description > Pitch Deck
Here's the first thread, where you can see it evolving over time.
1) First, generated title + body text w #gpt3
2) Added AI selected color
3) Added AI selected font
4) Added #dalle image
Here's the first thread, where you can see it evolving over time.
1) First, generated title + body text w #gpt3
2) Added AI selected color
3) Added AI selected font
4) Added #dalle image
👩🏫 Prompt > Lesson Slides (#education)
Here's "AutoLessons", for generating lesson slides.
Examples in thread.
Here's "AutoLessons", for generating lesson slides.
Examples in thread.
1/10 Working with LLMs is exciting & powerful, but there are few things I wish I had known getting started with prompt engineering.
Here are 8 tips for getting the most out of @Cohere's Generate, #ChatGPT, #GPT3, and other text generation models🧵👇🏻
🎬
Here are 8 tips for getting the most out of @Cohere's Generate, #ChatGPT, #GPT3, and other text generation models🧵👇🏻
🎬
2/10
I am using @CohereAI 's Generate to illustrate my points.
1. Start small!
Don’t try to do too much at once. Give the model simple, straightforward tasks and gradually build up to more complex ones.
A great example of a simple instruction can start with: "Write a ....".
I am using @CohereAI 's Generate to illustrate my points.
1. Start small!
Don’t try to do too much at once. Give the model simple, straightforward tasks and gradually build up to more complex ones.
A great example of a simple instruction can start with: "Write a ....".
3/10
2. Experiment and iterate.
Make a rule of experimenting with different versions of your prompt to see what works best for your model. Even small tweaks can do wonders for the quality of your output. 🧪👩🔬
2. Experiment and iterate.
Make a rule of experimenting with different versions of your prompt to see what works best for your model. Even small tweaks can do wonders for the quality of your output. 🧪👩🔬
POSSIBLE $2k Airdrop Guide
$Blur worth over $20K at ATH,
If you missed $Blur then don't miss this.
Don't wanna Cope? Participate in this Now.
A Thread 🧵. Like & RT For Frens
$Blur worth over $20K at ATH,
If you missed $Blur then don't miss this.
Don't wanna Cope? Participate in this Now.
A Thread 🧵. Like & RT For Frens
It's no news that testnet participants get nicely rewarded for having to participate in the #Testnet phase of @MyMetaTrader is the next-gen leverage trading platform built on @Arbitrum with #GPT3
@MyMetaTrader is the next-gen leverage trading platform built on @Arbitrum with #GPT3
Some of $MMT Use cases are;
Some of $MMT Use cases are;
looks like @openai bought the domain ai.com and hooked it up to chatgpt.
maybe everyone alr knew this and im just finding out lol...
accidentally found this out bc i was curious who owned the domain.
#ai #chatgpt #openai #gpt3
maybe everyone alr knew this and im just finding out lol...
accidentally found this out bc i was curious who owned the domain.
#ai #chatgpt #openai #gpt3
i think im playing with DALLE1 or something, this is kinda fun. it gives eight pictures and there are social share buttons embedded for each pic
With AskFred, we're applying all the incredible tech that powers #ChatGPT to a very tailored problem: Meetings!
@firefliesai is fortunate to collab with companies like @OpenAI
A special thank you to @vkhosla @SVenkatachalam @khoslaventures for making this possible
@firefliesai is fortunate to collab with companies like @OpenAI
A special thank you to @vkhosla @SVenkatachalam @khoslaventures for making this possible
Meetings is where we spend 1/3 of our work life
I now have an AI assistant that I can talk to about any conversation I've ever had... giving me perfect recall so I can better understand
My customers
My candidates
My colleagues
My company
I now have an AI assistant that I can talk to about any conversation I've ever had... giving me perfect recall so I can better understand
My customers
My candidates
My colleagues
My company
Knowledge originates in meetings
Knowledge is also buried in meetings
To address that, we made @firefliesai be the de-facto conversational knowledge base built on top of meetings
Tech like #gpt3 enables us to extract knowledge more effectively... #AskFred is a result of that
Knowledge is also buried in meetings
To address that, we made @firefliesai be the de-facto conversational knowledge base built on top of meetings
Tech like #gpt3 enables us to extract knowledge more effectively... #AskFred is a result of that
«Lo de automatizar cosas suena bien... pero no sé qué podría automatizar yo»
¿Te suena? Lo he escuchado mil veces y quiero ponerle remedio.
1 RT🔂 = 💡1 idea de proceso automatizable
🔥Vamos a ganarle la batalla al reloj🔥
¿Jugamos? Tengo 109 ideas preparadas
👇👇
¿Te suena? Lo he escuchado mil veces y quiero ponerle remedio.
1 RT🔂 = 💡1 idea de proceso automatizable
🔥Vamos a ganarle la batalla al reloj🔥
¿Jugamos? Tengo 109 ideas preparadas
👇👇
@salonsoweb 1.
¿Y si mañana Zuckerberg te cierra Instagram?
💡Idea de backup: cada vez que subes una foto, mándala automáticamente a Dropbox.
Ya que estás, guárdate también ese texto molón que has añadido para que no se pierda como lágrimas en la lluvia...
¿Y si mañana Zuckerberg te cierra Instagram?
💡Idea de backup: cada vez que subes una foto, mándala automáticamente a Dropbox.
Ya que estás, guárdate también ese texto molón que has añadido para que no se pierda como lágrimas en la lluvia...
@salonsoweb 2.
¿Y si pudieras leer un #podcast?
✅Detecta nuevos episodios
✅Descarga el audio desde el feed
✅Pásalo por una API de transcripción
✅Guárdate el resultado en un fichero de texto
(suena más loco de lo que es)
¿Y si pudieras leer un #podcast?
✅Detecta nuevos episodios
✅Descarga el audio desde el feed
✅Pásalo por una API de transcripción
✅Guárdate el resultado en un fichero de texto
(suena más loco de lo que es)
New WP: 🚨Artificial Intelligence Can Persuade Humans on Political Issues🚨Large language models (like #ChatGPT and #GPT3) are changing the world. Here we show that GPT3 can write messages that change people’s attitudes on important policy issues osf.io/stakv 🧵1/11
AI can now create art, write articles, and outperform humans on many tasks. It is an increasingly high-stakes concern whether AI could be used in political communication, such as online commenting, texting voters, and writing to legislators. shorturl.at/BCS59 2/11
But how effective is AI at influencing humans’ political attitudes? Research on political persuasion generally finds small effect sizes, and political persuasion is challenging, potentially drawing on complex skills like logical reasoning and clarity of expression. 3/11
.@OpenAI's #GPT3 now powering @todoist! 🔥
Here's a quick look at all the cool things you can do with it!
🔗 todoist.com/integrations/a…
Let's dive in! 🧵↓
Here's a quick look at all the cool things you can do with it!
🔗 todoist.com/integrations/a…
Let's dive in! 🧵↓
1/3
Break down tasks into smaller, more manageable actions
Break down tasks into smaller, more manageable actions
2/3
Making cryptic or complex tasks more actionable
Making cryptic or complex tasks more actionable
The following technique can be used to bypass AI text detection. Use it for ethical purposes only!👇
1. Generate whatever text you want with ChatGPT. Then reply to ChatGPT with the following:
"When it comes to writing content, two factors are crucial, "perplexity" and "burstiness." Perplexity measures the complexity of text. Separately, burstiness compares the variations👇
"When it comes to writing content, two factors are crucial, "perplexity" and "burstiness." Perplexity measures the complexity of text. Separately, burstiness compares the variations👇
of sentences. Humans tend to write with greater burstiness, for example, with some longer or complex sentences alongside shorter ones. AI sentences tend to be more uniform. 👇
We are getting closer to “Her” where conversation is the new interface.
Siri couldn’t do it, so I built an e-mail summarizing feature using #GPT3 and life-like #AI generated voice on iOS.
(🔈Audio on to be 🤯with voice realism!)
How did I do this? 👇
Siri couldn’t do it, so I built an e-mail summarizing feature using #GPT3 and life-like #AI generated voice on iOS.
(🔈Audio on to be 🤯with voice realism!)
How did I do this? 👇
How to get started with machine learning: From 0 to Engineer.
(thread)
(thread)
1. Math - via @khanacademy
Use these resources (all links in the last tweet):
- Khan Academy - Multivariable Calculus
- Khan Academy - Differential Equations
- Khan Academy - Linear Algebra
- Khan Academy - Statistics Probability
Use these resources (all links in the last tweet):
- Khan Academy - Multivariable Calculus
- Khan Academy - Differential Equations
- Khan Academy - Linear Algebra
- Khan Academy - Statistics Probability
2. Learn Python - via @freeCodeCamp
- Learn Python: Full Course For Beginners
- Intermediate Python: Programming Course
- Learn Python: Full Course For Beginners
- Intermediate Python: Programming Course
Okay, it's getting weird now...
Introducing "fuzzy.js" 🧶
An experimental approach to using #gpt3 as a back-end, communicated via HTML attributes.
Automatically fills in data.
Introducing "fuzzy.js" 🧶
An experimental approach to using #gpt3 as a back-end, communicated via HTML attributes.
Automatically fills in data.
You can get multiple items by repeating the same fuzzy_atr.
Just make sure the number of each match. (ie. if you have 3 "names", make sure have 3 "addresses")
Just make sure the number of each match. (ie. if you have 3 "names", make sure have 3 "addresses")
Attributes are used for describing the data field you want.
Describe the overall data you want when you call the fuzzyapi().
Describe the overall data you want when you call the fuzzyapi().
Wanna build the next ChatGPT?
Here's a 6 month learning plan for AI large language models (LLMs) from scratch.
(thread)
Here's a 6 month learning plan for AI large language models (LLMs) from scratch.
(thread)
1. Familiarize yourself with basic machine learning concepts and Python programming.
Books:
1.1. Introduction to Machine Learning with Python by Andreas Müller and Sarah Guido
1.2. Python Machine Learning by @rasbt
Books:
1.1. Introduction to Machine Learning with Python by Andreas Müller and Sarah Guido
1.2. Python Machine Learning by @rasbt
2. Learn about natural language processing (NLP) and its applications.
Books:
2.1. Speech and Language Processing by Daniel Jurafsky and James H. Martin
2.2. Natural Language Processing with Python by Steven Bird, Ewan Klein, and Edward Loper
Books:
2.1. Speech and Language Processing by Daniel Jurafsky and James H. Martin
2.2. Natural Language Processing with Python by Steven Bird, Ewan Klein, and Edward Loper
I always get a little excited when I convince #ChatGPT to do something it normally refuses to do 🏆
It wouldn't make me an email list of common names, but it would make a list of names with extra stuff added on the end on subsequent prompts 😉 #OpenAI #gpt3
It wouldn't make me an email list of common names, but it would make a list of names with extra stuff added on the end on subsequent prompts 😉 #OpenAI #gpt3
PS: None of those email addresses are real so far as I know. They are just different permutations based on input to ChatGPT 🤣 So if you think somehow you're going to reach any of those people you're dreaming. But you might find a scammer camped out on their name 🍿
However, one of these may or may not be real 😏
Oh wow!🥳
#gpt3 can build #nocode workflows with #ai.
Fed it two of my workflows as examples, and it successfully generated a third workflow based on query.
Imported to @zapier and it works! ⚡️
Right image: output. Left image: Imported flow.
#gpt3 can build #nocode workflows with #ai.
Fed it two of my workflows as examples, and it successfully generated a third workflow based on query.
Imported to @zapier and it works! ⚡️
Right image: output. Left image: Imported flow.
I should throw all my workflows as JSON in an airtable, and embed to find the most similar flows (cc @gpt_index), and use two of those as examples and this should get pretty robust.
Did some "testing" with the #ChatGPT (or #GPT3) detector by Hello-SimpleAI.
1. It's much more accurate than the GPT2 detector. All zero-shot prompt texts were detected. (but to be honest, anybody that works with ChatGPT on a daily basis can detect a ChatGPT text in 2 seconds)
1. It's much more accurate than the GPT2 detector. All zero-shot prompt texts were detected. (but to be honest, anybody that works with ChatGPT on a daily basis can detect a ChatGPT text in 2 seconds)
2. Long-prompt generations (especially from GPT3 playground) and summaries: not so much. About 8 out of 10 texts were not detected.
3. Reworked text or "text combinations" (several layers of AI interactions): zero detection
4. Translations in German: almost zero detection
3. Reworked text or "text combinations" (several layers of AI interactions): zero detection
4. Translations in German: almost zero detection
Tl;dr: Helpful tool for ChatGPT detection. But the bigger question is: should text (content) that is improved / re-written / corrected / translated / transformed by a #LLM be detectable? And if so, why don't we label highly altered or enhanced pictures, videos or audio?
It's been a blast building on #gpt3 since August. 😆
I've built and shared a few dozen apps and prototypes.
I often get asked this, so...
Here's a list of some favorites! 👇
I've built and shared a few dozen apps and prototypes.
I often get asked this, so...
Here's a list of some favorites! 👇
1) ⚡️Official @OpenAI x @zapier integration
This is the most popular by far.
Originally an unofficial integration I built for myself, it is now powering the next generation of #AI x #NoCode tools.
This is the most popular by far.
Originally an unofficial integration I built for myself, it is now powering the next generation of #AI x #NoCode tools.
2) 🔀Fuzzy Compiler
Turns "fuzzy/sloppy" code into clean code.
It speeds up my build a ton. It's mobile optimizes so I can generate code and add to Replit mobile app while I'm bouncing our baby down in the Ergo.
Turns "fuzzy/sloppy" code into clean code.
It speeds up my build a ton. It's mobile optimizes so I can generate code and add to Replit mobile app while I'm bouncing our baby down in the Ergo.
Free #AI "tip" for editorial teams & publishers (or other content-heavy businesses)
1. Create embeddings of ALL your content (every review, news, guide, video, podcast... for vector search). This can be done with @OpenAI but also with other models like @CohereAI or @huggingface
1. Create embeddings of ALL your content (every review, news, guide, video, podcast... for vector search). This can be done with @OpenAI but also with other models like @CohereAI or @huggingface
2. Train your own fine-tuned LLMs (again can be #GPT3 but can also be others like #BLOOM or Googles #FlanT5 or many others) for Q&A, Recommendations, Chat, and many more...
3. Connect your embeddings with your "own" LLMs and put a well-designed interface (text and voice) on your home pages and under every article or video you have.
