Everyday AI
Posts
When Will We Achieve AGI? One secret aspect holding us back.

When Will We Achieve AGI? One secret aspect holding us back.

Anthropic gets sued by authors, OpenAI partners with U.S. gov., future of AGI and more!

Everyday AI
August 20, 2024

👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI

Outsmart The Future

Today in Everyday AI
7 minute read

🎙 Daily Podcast Episode: When will we actually achieve AGI? We explain what AGI is and the one thing holding us back. Give it a listen.

🕵️‍♂️ Fresh Finds: Gemini API gets 1.5B tokens, Defcon raises $44M for military logistics and Microsoft updates its powerful Phi SLM. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: Anthropic gets sued for copyright, OpenAI’s first federal partnership, OpenAI releases fine-tuning for GPT-4o and Salesforce unveils new open-source models. For that and more, read on for Byte Sized News.

🚀 AI In 5: We show you how you can pull real-time data and visualize it inside Perplexity. Well, when it works. See it here

🧠 Learn & Leveraging AI: When will we reach AGI? And what does AGI even mean? We break down what you need to know. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about AMD's $49B acquisition, Google teasing new Gemini models, OpenAI SearchGPT waitlist and Trump's AI Taylor Swift controversy Check it here!

When Will We Achieve AGI? One secret aspect holding us back. 🧠

It's the trillion-dollar AI question.

When will we achieve Artificial General Intelligence? (And what the heck is it, anyway?)

We'll give you the 101 on what you need to know, and one secret that could be holding the official discovery back.

Join the conversation and ask Jordan questions on AGI here.

Also on the pod today:

• Evolution of AGI 🤖
• Definition of AI, AGI and ASI 💬
• Impacts of Advancements in AGI 🦾

It’ll be worth your 55 minutes:

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

Spotify | Apple Podcasts |
Google Podcasts | Amazon Music |

Upcoming Everyday AI Livestreams

Wednesday, August 20th at 7:30 am CST ⬇️

Transforming Global Supply Chains with AI

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight – MolyPix turns your sentences in to graphic designs, AgentQL provides easy data extraction and web automation and Fleak AI simplifies AI workflows for data teams.

Google - Gemini’s API now offers 1.5 billion free daily tokens along with some new tier pricing for the model.

We are giving developers 1,500,000,000 tokens for free everyday in the Gemini API
There is no stronger developer value proposition out there 🧵 (1/4)
— Logan Kilpatrick (@OfficialLoganK)
10:09 PM • Aug 19, 2024

AI Startup – Reliant’s product Tabular is helping to automate data extraction in research.

Money in AI – Defcon, an AI startup helping with military logistics, has closed a $44M seed.

Runway – Gen-3 Alpha’s image to video is now available on the iOS app.

LLM Updates — Microsoft has released Phi 3.5, an update to its small yet powerful language model.

Check This Out – This article discusses the big tech forces behind the AI PC push.

AI Video Updates — Luma has released a significant update to its AI video generator with Luma 1.5.

Dream Machine 1.5 is here 🎉 Now with higher-quality text-to-video, smarter understanding of your prompts, custom text rendering, and improved image-to-video! Level up. lumalabs.ai/dream-machine
#LumaDreamMachine
— Luma AI (@LumaLabsAI)
9:04 PM • Aug 19, 2024

1. Authors Sue Anthropic Over Copyright Concerns 🧑‍⚖️

In a significant legal development, a group of authors is suing Anthropic, alleging it engaged in “large-scale theft” by using pirated books to train its chatbot, Claude. This lawsuit adds to the growing number of copyright infringement cases against AI developers.

The authors contend that despite Anthropic's claims of being a responsible company, it profits from the creativity of writers without proper compensation.

2. OpenAI Partners with USAID to Enhance Efficiency 🇺🇸

OpenAI has officially partnered with the U.S. Agency for International Development (USAID), marking the agency as the first federal customer for ChatGPT Enterprise, aimed at reducing administrative burdens on staff. Under Administrator Samantha Power, USAID is prioritizing artificial intelligence to improve aid distribution and streamline operations.

Anna Makanju, OpenAI's VP, stated that the company is pursuing FedRAMP Moderate accreditation to securely manage sensitive federal data.

3. Salesforce Unveils Open-Source AI Models 🚀

Salesforce has announced the launch of xGen-MM (also known as BLIP-3), a suite of open-source AI models featuring an impressive 4 billion parameters.

These models excel at processing “interleaved data,” allowing them to seamlessly integrate text and images, which is crucial for applications like medical diagnostics and autonomous vehicles.

4. Meta Reports $1 Million Revenue Per Employee in 2024 🤑

Meta has achieved an impressive $1 million revenue per employee in the first half of 2024, bolstered by significant advancements in AI-driven advertising tools. Following job cuts in late 2022, the company has rebounded, with advertising still making up 98% of its revenues.

Utilizing tools like Advantage+, advertisers are experiencing a $371 return on every dollar spent, highlighting the effectiveness of Meta's AI strategies.

5. Hong Kong Drops New Guidelines for Generative AI 🇨🇳

The Hong Kong Monetary Authority just added new principles for using generative AI in consumer apps, focusing on governance and data protection. After a surge in banks adopting AI for customer interactions, HKMA's Alan Au is all about transparency and accountability in AI decisions.

The launch of the "GenAI Sandbox" is set to help banks explore innovative AI applications while keeping risks in check.

6. OpenAI Unveils Fine-Tuning Feature for GPT-4o 🛠️

OpenAI has announced the rollout of a new fine-tuning feature for its flagship model, GPT-4o, allowing businesses to leverage their own data for enhanced customization.

Set to launch today, this capability enables companies to tailor the AI for specific tasks, like creating an intelligent customer-service chatbot for skateboard inquiries. This allows organizations to customize the model with their own datasets—free training tokens are even available until September 23.

Previously, fine-tuning was not available on GPT-4o, but now organizations can take advantage of this powerful tool directly through OpenAI, potentially streamlining their AI integrations.

Perplexity's Real-Time Data Trick

Click Image To Play Video 👆

There’s some updated functionality inside Perplexity that no other LLM can do right now.

Perplexity can now pull real time data and visualize it.

We show you how to do it. Kinda. (It’s a bit finicky, TBH. But still pretty impressive.)

Check out today's AI in 5.

🦾How You Can Leverage:

It’s the (AI) elephant in the room.

The big bad wolf coming for our jobs.

The culprit?

AGI, or Artificial General Intelligence.

(Oh nooooooo!!!)

Here’s the reality — just about everything surrounding AGI is misunderstood.

Its origin.
Its always-changing definition.
It’s impending (or not?) arrival.

It’s the worst-kept secret in the AI space, but it’s a topic that we don’t talk about enough here on Everyday AI.

So, we went all in on AGI today.

And we didn’t just give our #HotTakeTuesday on when we think AGI would technically arrive, but uncovered some reasons why the powers-that-be may intentionally delay its announced arrival.

Ready to slap the big bad AGI wolf in the face?

POW
BAM
Other 80s Batman sounds

Here’s whathyu need to know about AGI.

Let’s get it, y’all. 👇

1 –AGI’s coming at us, SUPER fast 🏃

We're witnessing an AGI sprint that's leaving experts breathless.

You ever see this Ark Invest AGI predictions chart?

It's a wake-up call.

Five years ago, the brightest minds pegged AGI at 80 years out.

Today? A mere 8 years.

And we're betting that within a year, that prediction time range shrinks to 3 or less.

(Our personal take? AGI’s “official arrival” will be pretty undeniable in 5 years. In so many regards, we’re legit already there.)

This isn't just crystal ball gazing.

Tech titans like NVIDIA, Microsoft, Meta, and OpenAI are openly gunning for AGI.

It's no longer taboo - it's a race to the finish line.

NVIDIA's CEO predicts AGI within 5 years. Meta's crafting open-source AGI. And OpenAI? They're aiming to outperform humans at most economically valuable work.

2 – The goal posts are moving 🥅

Ready to make this even more confusing shorties?

Lolz.

Here's a mind-bender: by AGI definitions from 10-20 years ago, we might already be there. We unearthed some fascinating time capsules:

One old-school definition from 10ish years ago described AGI as essentially a system that can learn and solve problems across different areas, much like a human. Sound familiar?

Even Sam Altman, in his blogging days, painted an Aging picture that …. Kinda looks like today.

So why aren't we celebrating?

As AI advances, our expectations skyrocket. We're constantly redefining what counts as "general" intelligence.

Expect the AGI goal posts to continue to move.

Especially when…..

3 – The deal that could “delay” AGI 🤝

Stick with us here.

We call it #HotTakeTuesdays for a reason — we get to enter some opinionated takes on AI deveopment.

Here’s one — what benefit does OpenAI have to “achieve AGI”?

It’s literally part of its mission, sure.

But wouldn’t it be harder to fundraise trillions of dollars as a startup if you’ve….. completed your mission?

The OpenAI-Microsoft partnership is a complex dance that could be secretly slowing things down.

Microsoft has a $13 billion stake in OpenAI, and reported 49% equity.

Buuuuuuuuut here's the twist y’all – once OpenAI's board declares "AGI achieved!", that whole agreement reportedly changes.

Suddenly, Microsoft's stake in OpenAI's future AGI tech (reportedly) vanishes.

Think about it – OpenAI's getting crucial user data from Microsoft products, while Microsoft powers its tools with cutting-edge AI. Why upset that apple cart?

The plot thickens when you consider that OpenAI's GPT-4 is becoming the backbone of so many products. How do you separate "regular AI" from AGI when it's all coming from the same brain?

We sure don’t know.

But if we’re OpenAI, we’re not shouting ‘Mission Accomplished’ anytime soon, as it seems like there’s really no good reason to.

If you’re the Top AI Dog and fundraising like Facebook in the early 2000s, why would you change course and throw a monkey wrench in a good thing?

⌚

Numbers to watch

1.85 Billion Gallons

Virginia, one of the world’s largest concentration of data centers, water usage jumped by almost two-thirds between 2019 and 2023, from 1.13 billion gallons to 1.85 billion gallons.

Refer Friends to Win a Year of ChatGPT!

(Or your fave LLM like Claude, Gemini, Copilot, etc)

Now This …

Let us know your thoughts!

Vote to see live results

How many AI newsletters do you read?

Reply

or to participate.