Share this page!

Ethan Mollick Profile picture
Twitter logo
Dec 17 • 3 tweets • 2 min read
Entrepreneurship has been declining, even though 1/3 of Americans have had a startup idea in the past 5 years.

Part of the reason is that getting started can be daunting, and founders often need a little help to get moving. I think ChatGPT can do that. oneusefulthing.substack.com/p/chatgtp-is-m…
Folks are commenting that the reason is healthcare. Sure, but only half of Americans with ideas even do basic followup stuff, like search the internet for research. Overall, the road to entrepreneurship is long & bumpy and lots of people drop out at the first sign of difficulty.
And of course safety nets matter, too.

• • •

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
Dec 16
Asking ChatGPT riddles underlines how it can help us approach problems in a different (if not always right) way.

Example: “Turn me on my side and I am everything. Cut me in half and I am nothing. What am I?” The answer is usually 8 (turn it to become ∞, cut it to be zero).
It has no patience for clever riddles. This a practical problem, damn it.
It also solves the hardest logic puzzle in the world (more here: nautil.us/how-to-solve-t…) in a way that seems to be right & well-explained but which is actually nonsense.
Read 4 tweets
Ethan Mollick Profile picture
Dec 16
We don’t need Twitter anymore. I just had ChatGPT simulate it completely accurately.
Folks, I just made the simulation even more realistic.
It also perfectly simulates the experience of skimming through the Twitter Trending list in your 40s. Definitely read through the first couple.
Read 4 tweets
Ethan Mollick Profile picture
Dec 14
I wrote in Harvard Business Review about why I think AI has suddenly reached a tipping point for useful work & how that might shift what jobs look like in ways that are hard to anticipate.

I think every organization needs to grapple with what this means. hbr.org/2022/12/chatgp…
In case you haven't bought into the AI hype, I think you should. Most people aren't playing withe ChatGPT enough to see why it matters.

Here are four experiments (along with three bonus experiments) that you can do that might demonstrate its potential. oneusefulthing.substack.com/p/four-paths-t…
This Twitter thread shows you how to use ChatGPT to boost your writing. Very useful if you are an expert on a topic, it multiplies your abilities and time.
Read 4 tweets
Ethan Mollick Profile picture
Dec 12
Popular business advice is full of theories that are "useful, but not necessarily right," reducing complex issues into simpler factors, even if the framework is not particularly robust.

I regret to inform you that AI is very good at this. Witness the "Technology Adoption Matrix" ImageImage
I asked it for a theory of technology adoption, it tried to give me Rogers' Diffusion Theory, and then the Technology Adoption Cycle, but after enough prodding we have this new approach. Everything was done by the AI.

Here's how a consultant should pitch it, according to the AI. Image
It also wrote a book proposal. Image
Read 6 tweets
Ethan Mollick Profile picture
Dec 12
Business experiments help companies and startups succeed, but many of them do the wrong experiments because they test small changes, which can require huge samples if you are trying to identify tiny differences between two options

This paper on A/B tests at Bing shows why.... 1/
Most A/B results for mature products are small & precise (impacts of .06%) but some are much larger. To find those "fat tails," it is better to conduct lots of smaller, but less accurate, experiments. papers.ssrn.com/sol3/papers.cf…

Test & roll is one option. 2/
So it depends on the market:
Where tails are thin: “perform thorough prior screening of potential innovations & run a few high-powered precise experiments”
Where tails are thick: “run many small experiments, and test a large number of ideas in hopes of finding a big winner” 3/
Read 4 tweets
Ethan Mollick Profile picture
Dec 11
Teachers, the semester is ending, it is too late to change assignments, & almost all of your students are using AI. And while some are cheating, others use it as an editor/tutor

One option: ask they cite AI use & provide prompts. And assign a reflection on how AI helped (or not)
I did this, and it worked. It let students off the hook (no one knows what the plagiarism rules were for AI - just copying is wrong, but using AI to edit your essay? Suggest changes? Provide a draft?) & it helped me understand what was happening. Also reflections were thoughtful.
Next semester is going to be wild. oneusefulthing.substack.com/p/the-mechanic…
Read 5 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!

:(