Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Jeremy Nguyen ✍🏼 🚢 Profile picture
@JeremyNguyenPhD
Apr 10, 2023 • 7 tweets • 3 min read • Read on X
Scrolly
1/ ChatGPT has limited memory. It forgets details if your chat runs beyond ~8k words.

But Shogtongue lets you extend conversations by 5-10x easily

Here’s how: Image
2/ First: @gfodor's Shogtongue is ingenious.
It’s almost debasing it to use it for literal compression.

imagine: some genius harnessed microwave electromagnetic radiation
--now we use it to reheat day-old pizza Image
3/ After you've had a long chat,
ask GPT-4 to compress all of it
into its own weird language.

--where did it learn this language from?! Image
4/ Open a new GPT-4.
Give it the compressed message
and a bit of English context.

Shoggoth purists might give it only the shogtongue input.

But I like to help GPT out.

And there you go! You can now keep chatting for almost another ~8k tokens Image
5/ It’s still experimental.

-GPT-3.5 doesn’t work
-even GPT-4 via API struggles
-GPT-4 via ChatGPT is what works for me.

If it doesn’t work perfectly, that’s because we’re still discovering it right now. It’s exciting. Image
6/ PS. you can use it to have months-long
interactive journaling and review sessions

with learnings from months' worth of weekly reviews
—that’s what originally got me writing this thread.

Have fun. Let me know if you work out tweaks on the compression!
That's a wrap!

If you enjoyed this thread:

1. Follow me @RunGreatClasses for more of these
2. RT the tweet below to share this thread with your audience

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Jeremy Nguyen ✍🏼 🚢

Jeremy Nguyen ✍🏼 🚢 Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @JeremyNguyenPhD

Jeremy Nguyen ✍🏼 🚢 Profile picture
Mar 26, 2025
ChatGPT's new Image Generation dropped less than 24 hours ago

Here are 15 great examples of what you can do now, some limitations—and a hidden trick to get instant access if you're still waiting!

1. Life-like photos: Prompt: A lifelike gym photo showing Albert Einstein lifting dumbbells, wearing a tank top and gym shorts. Sweat glistens naturally on his forehead. The lighting comes from overhead fluorescents, casting real shadows. The background includes blurred gym equipment. The photo has a raw and authentic vibe, like a candid shot during a real workout session.
2/ Make a Greek statue of yourself:

Prompt: "Turn me into the old greek statue, background is full white" with 3 different angle photos
3/ Studio Gibli style portraits:

This is blowing up our feeds. Grant suggested sending your wife photos converted into studio gibli style images.

I can confirm that you can be a hero by making these for loved ones and relatives.
Read 17 tweets
Jeremy Nguyen ✍🏼 🚢 Profile picture
Apr 21, 2024
Meta's new AI—Llama3—beats an old version of GPT-4, Claude 2.1, and GPT-3.5!

You can run it:
- 100% free
- with 100% privacy (no data leaves your machine)

See the comments for an easy way to install 👇 Image
2/ GPT4All from @nomic_ai runs on Windows and Mac.

It's free. Other good options are @LMStudioAI or @ollama. But I'll go through running Llama 3 on GPT4All because it's also easy to chat with your own pdfs and documents with GPT4All (i.e. RAG).

Download it and install. Image
3/ You'll also need to download the model. (GPT4All is the interface. It's not the model).

Chose the new Llama 3 Instruct 8B. Image
Read 7 tweets
Jeremy Nguyen ✍🏼 🚢 Profile picture
Apr 17, 2024
MYSTERY SOLVED!

Why does ChatGPT use the word "delve" so much? We've seen a 10x increase in the proportion of medical studies using the word "delve" from 2022 to 2024. But why?

@alexhern at The Guardian might've just solved it. Thread below, complete with the trail of clues: Image
2/ The clues:

A few weeks ago, I tweeted a chart of the number of medical studies using the word "delve"—before and after ChatGPT came out (Nov 2022).
3/ Elon Musk commented.

Paul Graham shared his view.

Read 12 tweets
Jeremy Nguyen ✍🏼 🚢 Profile picture
Mar 31, 2024
What if ChatGPT had knowledge from your notes and documents—while keeping it all private?

You can run AI locally—no data leaves your laptop—for custom, secure answers from your "second brain."

It’s free.

Here’s how (more in comments):
2/ GPT4All by @nomic_ai makes it easy to run a completely local LLM informed by docs and notes you provide.

Download it from:
gpt4all [dot] io

Works on Windows and Mac. Image
@nomic_ai 3/ After you install GPT4All (the interface/ecosystem), you'll still need to download a model.

Choose a "chat-based" model, e.g. Nous Hermes 2 Mistral DPO.

If you've already downloaded models (eg in LM Studio), give GPT4All the directory where your model files are—it'll work. Image
Read 9 tweets
Jeremy Nguyen ✍🏼 🚢 Profile picture
Jan 9, 2024
Want a resume that opens doors?

Leah Graham fixed hers with *non-stupid* ChatGPT prompts. She got an interview at every single job she applied for.

Here’s how:
2/ Let’s start with the elephant in the room:

I’ll never insult your intelligence with those *stupid* GPT prompts littering Twitter:

—“Bruh, write me a resume that will get me a job as CEO at a Fortune 500!!!”.

No. That’s not how anything works. Image
3/ Leah’s prompts work. They delivered results.

She started by giving GPT context.

Then she did what most people don’t. She *did not* ask GPT to write her resume.

She told Jeep to ask her for more information. Image
Read 10 tweets
Jeremy Nguyen ✍🏼 🚢 Profile picture
Jan 8, 2024
What if ChatGPT had knowledge from your notes and documents—while keeping it all private?

You can run AI locally—no data leaves your laptop—for custom, secure answers from your "second brain."

It’s free.

Here’s how:
2/ GPT4All by @nomic_ai makes it easy to run a completely local LLM informed by docs and notes you provide.

Download it from:
gpt4all [dot] io

Works on Windows and Mac. Image
@nomic_ai 3/ After you install GPT4All (the interface/ecosystem), you'll still need to download a model.

Choose a "chat-based" model, e.g. Mistral OpenOrca.

If you've already downloaded models (eg in LM Studio), give GPT4All the directory where your model files already are—it'll work. Image
Read 8 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

Send Email!

:(