Post

How to get URL link on X (Twitter) App

On the Twitter thread, click on or icon on the bottom
Click again on or Share Via icon
Click on Copy Link to Tweet
Paste it above and click "Unroll Thread"!
More info at Twitter Help

Aadit Sheth

May 17, 2023 • 11 tweets • 4 min read • Read on X

AI is more dangerous than you think.

Recently the founder of ChatGPT, Sam Altman, talked about AI regulation.

The dangers, risks, and precautions to take.

The recording is 4h 38mins.

Here are 7 most important things you need to know:

1. Opening statement

AI has the potential to improve nearly every aspect of people's lives but it also creates serious risks.

2. Misinformation risk

Companies should review the test results before releasing the model.

Independent audits are also necessary to ensure the model's accuracy and reliability.

3. Job risk

The jobs of tomorrow will be even better than the jobs of today.

New technologies will create new opportunities.

The jobs of today will be enhanced by AI.

4. Sam's worst fear

"My worst fears are that we, the field the technology the industry, cause significant harm to the world."

5. Election misinformation

There is concern about the impact of AI on elections.

The industry and government should work together to address this.

Policies and monitoring are in place to detect AI-generated tweets.

6. Best ways to regulate AI

• Create a new agency to license & regulate AI models.

• Create safety standards for AI models that identify & mitigate dangerous capabilities.

• Require independent audits from experts to ensure compliance with safety standards.

7. On language inclusivity in LLMs

It is important to make models available in many languages.

"We worked with Iceland and will partner with others to make models available in lower-resource languages."

If you enjoyed this thread, you can subscribe to my newsletter, Prompts Daily for more AI insights, prompts and tools to keep you ahead:

neatprompts.com/subscribe?utm_…

Enjoyed this thread?

Subscribe to the newsletter for more AI insights, prompts and tools to help you become a machine at work: neatprompts.com/subscribe

https://twitter.com/aaditsh/status/1658898874554257408

If you found this valuable, feel free like/share with your friends 🙏

https://twitter.com/aaditsh/status/1658898874554257408

• • •

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

This Thread may be Removed Anytime!

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

More from @aaditsh

Aadit Sheth

@aaditsh

Jul 26

human skills will become MORE valuation in an AI-driven world.

look. everyone is talking about ai replacing jobs. they're missing the point.

ai isn't making human skills obsolete. it's making certain human skills more valuable than ever:

1. emotional intelligence: ai can analyze data, but it can't read a room. your ability to understand and manage emotions becomes your superpower.

2. creativity: ai can generate, but it can't innovate. true creativity - the kind that leads to breakthrough ideas - remains uniquely human.

Read 7 tweets

Aadit Sheth

@aaditsh

Jul 25

ever tried to cut down a tree with a butter knife?

that's what using ai with a bad prompt feels like.

here's why sharpening your ai 'axe' is the game-changer you didn't know you needed:

in the world of ai, your prompt is your axe. and boy, does it need sharpening.

think about it:

a dull axe: "write a blog post about ai"

a sharp axe: "write a 500-word blog post about how ai is changing customer service in the banking industry, focusing on chatbots and predictive analytics"

see the difference?

here's why sharpening matters:

• garbage in, garbage out: vague prompts get vague results
• efficiency: a sharp prompt saves time. lots of it.
• consistency: well-crafted prompts give reliable outputs
• creativity: paradoxically, constraints breed creativity. sharper prompts, more creative ai

Read 5 tweets

Aadit Sheth

@aaditsh

Jul 7

I recently started using Claude much more.

More than GPT-4o.

People really need to try different models to see what works for them.

Here are 7 amazing use cases that will inspire you to try something different than GPT-4o:

1. Convert any pitch deck into an interactive dashboard for investors.

2. Transform a simple PDF into an interactive dashboard.

Read 10 tweets

Aadit Sheth

@aaditsh

Jun 30

2 really smart friends considered canceling their ChatGPT subscription last week.

Why?

Claude 3.5 Sonnet.

It does many things better than GPT-4o.

Here’s everything you need to know:

Everything began on June 20, 2024, when Anthropic announced their most efficient model, Claude 3.5 Sonnet.

Since then, content creators on X have been enthusiastically testing various experiments.

After they released the benchmarks, Sonnet completely defeated GPT-4o

Read 14 tweets

Aadit Sheth

@aaditsh

Jun 24

🚨Breaking: Anthropic dropped Claude 3.5 Sonnet.

It's the most powerful and affordable model yet.

People are already doing incredible things with it.

7 mind-blowing use cases: 👇

First, check their official release:

Their vision model is impressive.

Read 14 tweets

Aadit Sheth

@aaditsh

Jun 6

Perplexity just launched a new feature called Perplexity Pages.

It acts like a content assistant, saving you time and effort.

Here's how you can utilize this:

You’ve used Perplexity to search for answers, explore new topics, and expand your knowledge.

Now, it’s time to share what you learned.

It’s like having an AI-powered Wikipedia at your fingertips.

Here's the step-by-step process:

Step 1: Go to Perplexity
Step 2: Click on "+"
Step 3: Click on "Page"
Step 4: Provide your topic
Step 5: Click on "Convert to Page" and share it with others