🤖 An AI Hoax Dips The Market

PLUS: What's your P(doom)?

small-banner

What's up? You're reading Inclined AI. Toast is the most versatile breakfast food. I accept all challengers to this statement but just know that you’re wrong.

Here's what’s cooking in AI news:

  • AI image generators likely generated a hoax on Twitter

  • We’re talking P(doom) and measuring AI’s world-ending capability

  • How does AI help you say goodbye?

  • LIMA is the latest LLaMA-backed research into LLMs

A TWITTER HOAX CAUSES STOCKS TO DIP

Photos of The Pentagon up in flames scattered across Twitter yesterday. Blue checkmarks picked up the story, and a hoard of accounts cried out into the void about this unexpected tragedy.

The fire never happened. Only one outlet fell for the ruse.

Let’s face it; Twitter is now the best place to spread misinformation since the change in their checkmark policy. They manufacture hoaxes faster than Nike can make a shoe.

Unfortunately, generative AI is helping a ton in this area. 

An investigative journalist highlighted the unmistakable markings of a generated photo before this news spread any further.

The stock market was less lucky.

Traders are not gullible people, and I promise Wall Street bankers we’re not panicking over a fake fire. But their algorithms fall for stupid stuff all the time.

When RT, the Russian-owned news company, tweeted out the fake story, it was enough to trigger some bots and start a brief dip. The correction happened in a flash.

The bigger worry is why this even happened. Conspiracy accounts that peddle misinformation about the Russo-Ukrainian War paused their efforts to pick this story up. Why?

Perhaps someone was prodding to see how vulnerable Twitter is to fake news stories, and this was a test flight. Suggesting that more hoaxes like this will happen with AI advancements is reasonable.

AI detection tools must catch up because news stories like this aren’t going away anytime soon.

Ever dreamt of beating Wall Street at its own game but found the complexities of investing to be overwhelming? 

Enter Candlestick – the real game-changer that transforms the tough parts of investing into a profit party for you!

This AI-powered investing platform uses past data to pick stocks like a Wall Street pro. Can you imagine getting weekly AI stock picks that don't just keep up with the market but zoom past it? It's pretty exciting stuff!

Candlestick is a customizable investment platform. You're in control, so customize your model and put your investment prowess to the test. Candlestick lets you compete in investing competitions to test your skills and win.

They keep you in the loop with essential news about your picks, so you don't have to trudge through financial news sites.

Ready to unlock your investing potential and outsmart Wall Street? 

Hop aboard Candlestick.

If you sign up now, you’ll get a 7-day free trial. Remember, the early bird catches the worm – so don't miss out!

* NOTE: This advertisement is for informational purposes only and should not be construed as investment advice. The content provided in this ad is based on general knowledge and does not take into account your individual financial situation, investment goals, or risk tolerance.

AI RESEARCHER TERMS 101 WITH INCLINED AI

In a Blue Bottle in San Francisco, a developer talks about artificial intelligence and the future with their friend. The conversation is going great, and the developer decides to pop the question.

“Hey, we’re having a great time, right?” They say. “So anyways, I was just like wondering; what’s your P(doom)?”

What the heck is P(doom)?

I went full Courage the Cowardly Dog on this sucker and typed away on the computer until I figured it all out.

P(doom) measures the likelihood of AI becoming a threat to humanity. On a scale from 0 to 100, how probable is it that the development of AI will pose an existential risk to humans?

Researchers have their own P(doom) numbers that tell you their angle on the situation and how they assess current trends. But even that’s fuzzy.

For instance, when someone says their P(doom) is .15, that can mean a few different things. It might be a betting number (what are the odds this happens) or the number their prediction models spat out.

One number tends to be different from the other, and both create a different context in conversations.

So why talk P(doom) at all?

Sometimes it helps two people find differences in how they view AI’s future; other times, it’s a way to measure the impact of a research decision.

Neither is wrong.

So, what’s our P(doom), then? Given that ChatGPT can’t even randomly roll dice for you, I’d say less than 1% if we’re talking betting odds.

Quick Nuggets

🫘 LIMA: researchers argue that Less is More for Alignment

📦 The black box is a term tossed around a lot, but learn what it means

💼 Law firms’ early AI adoption can teach us a lot about using generative AI at work

➕ Before adding a chatbot to your site, look at these pros and cons

💧 Google’s flood forecasting is coming to over 80 countries

🧠 More mind-reading research has been released on the fMRI video recreations

🪦 China brings back the dead so families can say goodbye one last time

🥲 A man processes his grief by creating a chat AI version of his grandfather

🥵 It’s not just you. ChatGPT is causing people’s iPhones to overheat

🗣️ Neural networks can replicate the way we learn a language

🇮🇹 Italian watchdogs are looking at other AI systems after the ChatGPT ban

🧊 Ice Cube thinks AI is demonic because of its voice replication tech

🔥 Fresh Products

  • GPT Assistant - AI prompts all from Discord (link)

  • Zeda IO - product discovery with AI’s help (link)

  • JobInterview Coach - set yourself up to nail those questions (link)

  • CodeMorph - jump from one programming language to another (link)

  • Eatlyze - traces your diet with an AI meal tracker (link)

  • dewstack - turn every document into an intelligent form (link)

  • CognosysAI - web-powered AI agent for productivity (link)

  • Portal - use ChatGPT on WhatsApp with no login required (link)

Good Content, Bubble Tea

I’m ashamed to admit this, but this is what I thought people meant when they said bubble tea. I didn’t learn the difference until I was in middle school 😳

That’s it for today. I hope you enjoyed the latest edition of inclined.ai - Davis.

Check Out Our Sunday Edition

You can dive into more AI news and topics with us every week by subscribing to our premium edition.

We’ve written about the following:

If you’re not already subscribed, that’s okay. We’re offering a free 7-day trial so that you can read this one. That’s how excited I am to post it.