Share this page!

Enter Twitter Thread URL to Unroll

Needs to be the full URL like: https://twitter.com/threadreaderapp/status/1644127596119195649

How to get URL link on 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
Ethan Mollick Profile picture
@emollick
May 3 • 5 tweets • 5 min read Twitter logo Read on Twitter
Scrolly
The new version of Midjourney that released yesterday shows how far AI has come in making commerical-level images from text alone

Here is what you get for "modern outfits inspired by Van Gogh/ Basquiat/ Monet/ Rothko, fashion photoshoot" Each one is the first try, no revisions. ImageImageImageImage
Modern outfits inspired by...
Escher
Dali (honestly, everything it created based on Dali was interesting)
Turner
Rembrandt ImageImageImageImage
More Dali ImageImageImageImage
Kahlo
Munch
O'Keefe
Klimt (everything based on Klimt is amazing) ImageImageImageImage
Klimt (yes, again)
Koons
Magritte
Seurat ImageImageImageImage

• • •

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

Keep Current with Ethan Mollick

Ethan Mollick 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 @emollick

Ethan Mollick Profile picture
May 2
So it looks like ChatGPT with Code Interpreter can do basically any visualization technique you ask. And you can tell it to make charts "more beautiful"

(I actually asked it to generate artificial data to create the charts to show what they could do, which the AI happily did😬) ImageImageImageImage
And yes, you could do this all in Python. But I don't know Python, and doing this stuff in any coding language is always a pain (I have STATA nightmares to prove it). And it suggests data, cleans the data, fixes graph issues, etc. In seconds. Image
Name any visualization technique, no matter how obscure. Lets test it!
Read 4 tweets
Ethan Mollick Profile picture
May 1
Academia is in for a wild ride

>Upload census data & data dictionary into GPT with Code Interpreter
>"I would like you to generate some interesting draft hypotheses about industries & MSAs and then to test them with the data. make assumptions if you need to"
>"Put it in a paper" ImageImageImage
It is obviously not a top journal paper, or anything close. But this took me less than 10 minutes, and I did not do any work to find an interesting dataset or guide the AI in any way.

If nothing else, academic publishing is about to be overwhelmed.
I wouldn't take a lot of comfort in the fact that it just did OLS in the example... Image
Read 5 tweets
Ethan Mollick Profile picture
Apr 29
Overrated & underrated forms of GPT compared to online hype

Overrated: AutoGPT (awesome future possibilities, doesn’t work well yet), ChatGPT plugins (ditto), ChatGPT with web (currently struggles)

Underrated: Bing Creative Mode (GPT-4 & internet), ChatGPT with code interpreter
ChatGPT with web access
AutoGPT
Read 5 tweets
Ethan Mollick Profile picture
Apr 23
Another OMG GPT moment.

I fed it a business planning spreadsheet used only in Wharton classes (no answer key). When I just pasted in the Excel text, the AI read & understood it.

And it identified the key assumptions (a point in class)

And it conducted a sensitivity analysis! ImageImageImageImage
Since there were some questions, I literally just copied and pasted the Excel data into ChatGPT👇, without formatting it or uploading it in any way.

I have access to plugins, which let ChatGPT use Wolfram for the sensitivity analysis, but otherwise it didn't use any plugins. Image
Its ability to work with data is surprising.

I pasted in the raw data from a private VC spreadsheet tracking opportunities that I was given a decade ago. It figured out the data structure, and, once it was told it was data from a VC, correctly understood the details and outcomes ImageImageImage
Read 4 tweets
Ethan Mollick Profile picture
Apr 21
This was crazy.

I asked Bing to create lessons on entropy for middle schoolers (more, and prompts: oneusefulthing.org/p/democratizin…)

I also asked it make a game to explain it. It gave me working code for a game it designed, where I can change entropy and see the results on molecules.
Code it created, for a game it designed. I literally did nothing but ask: "give me code for a small game that teaches a concept related to entropy for middle schoolers"

# Import Pygame library
import pygame
import random
import math

# Initialize Pygame
pygame.init()

# Define… twitter.com/i/web/status/1…
Also, I can’t code. I just did what the AI told me to to get Python working.
Read 4 tweets
Ethan Mollick Profile picture
Apr 20
Why I don't take much comfort in "GPT got this wrong, so AI will never be able to do this thing right," in one graph. Image
I pulled this all from the OpenAI whitepaper: openai.com/research/gpt-4
And yes, contaminated data is an issue, but recent tests with credibly uncontaminated data show similar results.
Read 4 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 on Twitter!

Send Email!

:(