AGI by 2027, Apple & OpenAI, and GPT's Multiple Outages
PLUS HOT AI Tools & Tutorials
Welcome to our weekly digest!
As we count down the days until WWDC 2024 (hey, three days to go), OpenAI continues to get caught up in scandals, startups are still raising big money, more AI tools are popping up worldwide, and tech companies integrate chatbots into their products. And it's all waiting for us to discuss.
So let's talk about it!
This Creators’ AI Edition:
Featured Materials 🎟️
News of the week 🌍
Useful tools ⚒️
Weekly Guides 📕
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
(Bonus) Materials 🎁
Featured Material 🎟️
Leopold Aschenbrenner & The Largest AI Forecast
This story consists of two independent (yet very connected) elements. Therefore, we will divide it into two parts. First, we will discuss the underhanded games in OpenAI, and then talk about the predictions of the main person in this news.
Do you remember Leopold Aschenbrenner? I guess not, so let me explain. A few months ago, The Information published a report claiming that OpenAI fired Aschenbrenner, a member of the Superalignment team, for disclosing confidential information and possibly being involved in the brief ousting of Sam Altman.
The story had an unexpected sequel and many interesting details two months later.
Aschenbrenner appeared on Dwarkesh Patel's podcast and gave his version. In his conversation with Patel, he stated that he did not divulge OpenAI's secrets but cared a lot about the company's security. Aschenbrenner said he found several vulnerabilities that could have led to the theft of secrets from foreign actors. So, he tried to alert management to the potential problems. (according to him)
He was ignored by the executives, so he went to the board and was ignored again. Then, Aschenbrenner got a pre-warning from the HR department and was fired a few months later. The company stated the incident was a contributing factor.
Interestingly, Aschenbrenner said that “the incident” that preceded the firing was innocuous. He had written a document on “preparedness, safety, and security measures,” which he shared with some external researchers. According to the researcher, this was “totally normal at OpenAI at the time,” and the document had been scrubbed of sensitive information.
Keep your mailbox updated with key knowledge & news from the AI industry
Aschenbrenner suggests that the company was looking for an excuse to fire him because he didn't sign a letter to bring Altman back to OpenAI.
There's a lot of confusing stuff happening here, and we don't have enough facts to make the puzzle fit together. Therefore, I urge you not to jump to conclusions and to treat the statements of both sides with a reasonable degree of skepticism.
More importantly, getting fired or his point of view isn’t the main reason why Aschenbrenner appeared in featured material. The main reason is below.
AGI by 2027?
Concurrent with his podcast visit, a former OpenAI employee published a 165-page document of predictions based on his experience and knowledge of AI. At the beginning of this file, he thanked his colleagues, including key people from OpenAI.
Obviously, we can't discuss every topic of this huge document here, so let's focus on the main points. If you want to learn more, you can read the full version at this link.
Highlights from Aschenbrenner's forecast:
Aschenbrenner predicts that AGI (Artificial General Intelligence) could be created as early as 2027. AGI is AI that matches human intelligence in its intellectual abilities.
This also overlaps with recent predictions by Jensen Huang and Sam Altman.
According to Aschenbrenner, power consumption is gradually becoming a constraint (AI will consume 1 GW by 2026 and about 10 GW by 2028) due to a lack of spare capacity.
According to Aschenbrenner, companies are willing to invest trillions of dollars to develop the computing capacity needed to build AGI. By 2025-2026, these machines could surpass the intelligence of college graduates and any human being by the decade's end.
Microsoft is already building a similar cluster for OpenAI.
Aschenbrenner warns of a possible AI race between the U.S. and China that could lead to an all-out war in the worst-case scenario. He also draws attention to the "data wall" problem of limited data on the internet for training AI models.
By 2030, annual investment in AI will reach $8 trillion dollars, and AGI will be just the beginning. Once it is created, the transition to ASI (super intelligence) will occur almost immediately. AGI will be so smart that it can self-learn, which will happen quickly.
If we were talking about the words of some insider, it could be written off as an overblown fantasy. However, given that Aschenbrenner has specific knowledge and understanding of the processes inside OpenAI, his predictions look impressive.
News Of The Week 🌍
Raspberry Pi Partners with Hailo to Create AI Extension Kit
The Raspberry Pi Foundation has teamed up with startup Hailo (which recently raised $120M) to create a new small single-board computer with AI support. The AI Kit offers an accessible way to integrate local, high-performance, power-efficient inferencing into various applications. You can find all the technical details (and quite a few) here.
The latest AI Kit is already available for $70.
I don't think the Raspberry Pi needs much introduction. When we talk about the democratization and accessibility of modern technology, we bring this thing up one way or another (And if we don't, we definitely should). And now the Raspberry Pi promises to democratize hardware AI development. That sounds pretty cool.
ChatGPT Faced Multiple Outages
A few days ago, on June 4, an apocalypse came to the AI world. But it lasted for about three hours. ChatGPT encountered numerous outages, and users worldwide could not access the chatbot. OpenAI had to work on these for hours to make the service available to users again.
The problem has now been fixed, and ChatGPT is back up and running.
Yes, you may say that since there are no more outages, this event is not worth mentioning. But I just wanted to point out the panic that continued in X and Threads for several hours. It also makes you think about how we're falling into dependence on neural networks.
GetWhy Raises $34.5M for Its AI Research Platform
GetWhy, based in Copenhagen, has joined the long list of AI startups receiving big investments. The company promises to change the consumer research market by using a proprietary platform to collect and analyze consumer brand information on a large scale. GetWhy's platform allows its clients to conduct qualitative research by extracting information from video content.
GetWhy says their product has the potential to make analytics and research much cheaper. And that may well be true. The startup's services are already being used by Nike, Heineken, Coty, The Coca-Cola Company, Adidas, and Carlsberg Group.
Sharing is caring! Refer someone who recently started a learning journey in AI. Make them more productive and earn rewards!
Meta Adds AI-Powered Features to WhatsApp Business App
Speaking at the Meta Conversations conference in Brazil, Mark Zuckerberg announced new AI tools for entrepreneurs using WhatsApp Business. The app will soon include features for quickly creating ads on WhatsApp and generating answers to popular customer questions. Meta's CEO says the company wants to create more consumer-focused AI agents.
An expected update that sounds pretty useful. But I have concerns about how such a decision might hurt competition among AI managers. If Meta monopolizes this kind of functionality in its apps, it could really hurt a lot of startups.
Apple Will Announce Partnership with OpenAI
We are counting the days and hours until the WWDC 2024. According to Bloomberg, we’ll finally get official confirmation of the agreement between Apple and OpenAI during this conference. It is expected that the update may affect Siri. I'd like to believe that Apple's voice assistant will finally be more valuable than a timer tool (does anyone use it otherwise?).
Save the date: June 10, 2024.
Columnist and one of the top insiders on the Apple products, Mark Gurman, has quite accurately noted that the the agreement will be a key focus of Apple’s сonference next week and it shows how much the power in Silicon Valley has shifted over the past few years. I don't think there's much to add.
Nvidia Passes Apple in Market Cap
Following Mark Gurman's commentary on changes in Silicon Valley. Here's what gives a focus on attractive technology. Nvidia is now the second-most valuable public company. The company has surpassed Apple and is now second only to one IT giant, Microsoft. In addition, Nvidia also hit a $3 trillion market cap milestone after shares rose over 5%.
According to CNBC, investors also feel more confident that Nvidia's massive growth in sales to a handful of cloud companies can persist. In the most recent quarter, revenue from its data center business, including sales of GPUs (which are heavily used for AI training), rose 427% from a year earlier to $22.6 billion. That's about 86% of the company's total sales.
If you want to read a little more about how AI has booed Nvidia, we have a newsletter for you:
Share this post with friends, especially those interested in AI stories!
Useful Tools ⚒️
Databutton – Let AI build your next SaaS application
Falcon AI – Your AI Scrum Master
Exante – AI-powered contract repository
Deepshot – AI lip syncing & generation with no limits.
Fliki – Turn text into videos with AI voices
If you're looking for a simple way to create videos, you can check out Fliki AI. It's one more text-to-video platform that makes content creation easy by converting text into videos with AI-generated voices. It supports over 75 languages and 1300+ voices, making it easy to reach a global audience. Plus, it's excellent for creating ads, tutorials, and even podcasts without needing any technical skills.
Weekly Guides 📕
Invideo AI Tutorial: Best AI Video Generator in 2024
RIP ANIMATORS! Animate ANYTHING With AI IN 1-CLICK!
How To Access GPT Store For Free: ChatGPT-4o is OpenAI's Newest Flagship Model
13 Amazing ChatGPT-4o Shortcuts for iPhone
How to add AI Chat GPT in Excel Ad-ins | How to Use AI in Excel
AI Meme Of The Week 🤡
AI Tweet Of The Week
(Bonus) Materials 🏆
Your Right to Repair AI Systems | Rumman Chowdhury | TED
Prepare to Get Manipulated by Emotionally Expressive Chatbots
Sam Altman Was Bending the World to His Will Long Before OpenAI
How Apple Fell Behind in the AI Arms Race
Better AI Models, Better Startups
How was this week in AI? Share your content and ideas in the comments to this post so we can discuss or include them in the next edition!