- Not A Bot
- Posts
- 🏟️ Apple enters the arena
🏟️ Apple enters the arena
PLUS: Google Meet is testing AI-generated backgrounds.
New here? Welcome to NAB, and make sure to subscribe for more!
Happy Thursday, fellow humans. 👋
Yesterday was Meta’s big day. Today is Apple’s. Big Tech is taking turns launching big AI news, and we’re here for it.
Let’s dive into it…
Have anything you’d like to share with over 40k subscribers?
🧵 In today's edition:
🏟️ Apple enters the arena
🧑🎓 ChatGPT passes Harvard Freshman Year
📰 OpenAI aims to help journalism
🤑 AI Fundraising News
🤖 Top AI News
🏟️ Apple enters the arena
After barely mentioning AI at WWDC this past June, Apple revealed that they have been developing their own framework for creating large language models, which could compete with those of OpenAI and Google.
The framework, called "Ajax," has also been used to create “Apple GPT,” a tool pre-emptively launched late last year for internal use only but later halted over security concerns.
This is great news for Apple for a number of reasons.
For one, Siri could finally get the upgrade it deserves. Despite launching in 2011, way before other competition, Siri has taken a back seat as “Apple focused on other areas and adopted fewer features in favor of privacy.”
This could change for the better if Apple can utilize its framework to make Siri a go-to for all iPhone users.
Secondly, Apple had been keen on commercializing AI but couldn’t find the most effective way to do it.
Taken from Apple’s 2022 Annual Report: Bloomberg
Now, they may have a way to do so responsibly.
Add this to their planned launches for Quartz, a new health coaching service that uses AI to personalize plans, and a future electric car, and Apple could become a mainstay in the Trillion Dollar market cap club.
Exciting times ahead for Tim Cook and his gang.
Read more: Bloomberg
🧑🎓 ChatGPT passes Harvard Freshman Year
Credits: Maya Bodnick // Carly Bodnick
ChatGPT continues to impress my parents.
A Harvard student (Maya Bodnick) conducted an experiment to test ChatGPT's ability to pass a typical freshman year at Harvard for someone in a Humanities major.
The results? A. A. A. A-. B. B-. C. That roughly translates to a 3.34 GPA. Wild.
The experiment was conducted by asking eight professors to grade ChatGPT's essays generated in response to real Harvard prompts.
While some feedback included:
“It is beautifully written!”
“Well-written and well-articulated paper.”
“Clear and vividly written.” “The writer's voice comes through very clearly."
Other feedback came in more critically, calling out the essays' verbosity and use of incorrect facts to make political points (a common problem with most chatbots today).
Overall: The rise and popularity of generative AI has rocked the higher education sector and will eventually cause us to rethink how we teach certain subjects, especially pertaining to the social sciences.
Read more: Yahoo Finance | Original Blog Post (great read!)
📰 OpenAI aims to help journalism
OpenAI, heck AI in general, is on its way to taking over the world.
And Sam Altman, the revolution's leader, has spent much time convincing world leaders about the importance of this next decade in AI.
To help compliment his efforts, OpenAI has partnered with the American Journalism Project, a venture philanthropy organization that provides grants to non-profit newsrooms, pledging $5 million to fund various AI projects within newsrooms.
AJP plans to use the money to:
Build a “technology and AI studio” staffed by a team of experts to share ways for local newsrooms to benefit from AI
Use the remaining funds to disperse direct grants to partner organizations that can be used to set up AI pilot programs.
To the naked eye, this seems like a great initiative to help newsrooms across the country take advantage of a technology that can clearly help them.
But if we are to dig a little deeper, this seems like a way to get the media on OpenAI’s side while getting access to vetted information they could use to train their models.
This could be a follow-on effect from their deal with Associated Press - an arrangement that sees OpenAI licensing “part of AP’s text archive” as AP “examines potential use cases for generative AI in new products and services.”
All of this seems like lawyer talk for “give me more data.”
Read more: Gizmodo
🤑 AI Fundraising News
Preply, the language app known for its live tutors, closes out Series C at $120M and doubles down on AI.
FedML raises a $11.5M seed round to combine MLOps tools with a decentralized AI compute network.
🗞️ AI Quick-Bytes
What else is going on?
Google Meet is testing AI-generated backgrounds.
AI prodigy warns lawmakers on China’s ambitions: ‘AI is China’s Apollo Project.’
Sberbank CEO tells Putin of huge returns on its AI investments
CodeSee adds generative AI to explore code bases with natural language queries
Nothing CEO Carl Pei on the Phone 2, AI, and the future of gadgets
China’s Tech Overseer Promises to Back AI Computing Push
🐦️ Tweet of the day
Interesting data + insights here 👇️
Is GPT-4 worse now? 👀
New research says yes.
Here're how it worsened:
Finding prime numbers? Down from 98% to 2.4%.
Writing good code? Down from 52% to just 10%.
Reasoning tasks? More mistakes now.Wonder why? Here are a few guesses:
- OpenAI might be aiming for speed and… twitter.com/i/web/status/1…
— anita (@anitakirkovska)
5:35 PM • Jul 19, 2023
What did you think of today's newsletter? |
And that does it for today's issue.
As always, thanks for reading. Have a great day, and see you next time! ✌️
— Haroon: (definitely) Not A Robot and @haroonchoudery on Twitter
P.S.: Check us out on our socials - snacks on us if you also hit follow :)
Reply