We had two such anti-technology jobs moral panics in the last 20 years — “outsourcing” enabled by the Internet in the 2000’s, and “robots” in the 2010’s. The result was the best national and global economy in human history in pre-COVID 2019, with the most jobs at the highest wages ever.
Greetings, carbon-based lifeforms and advanced AI constructs!
It’s time for another edition of our AI newsletter – featuring all the latest news and analysis on the rise of the machines.
- Marc Andreessen on why AI won’t put us all out of a job
- Brave Browser Launches AI Summarization
- 200-Year-Old Math Opens Up AI’s Mysterious Black Box
- The inside story of how ChatGPT was built from the people who made it
- Is OpenAI making money from ChatGPT API?
- Stocks like Google, Microsoft, Amazon and Nvidia are likely to see an impact from OpenAI.
Why AI Won’t Cause Unemployment
Marc Andreessen is seen as a visionary in Silicon Valley and has been named one of Time Magazine’s 100 most influential people in the world.
He recently started a substack and one of his pieces is debunking the fact that AI will put us all out of jobs.
He points out that fears about a new technology causing massive unemployment is nothing new.
We had two such anti-technology jobs moral panics in the last 20 years — “outsourcing” enabled by the Internet in the 2000’s, and “robots” in the 2010’s.
The result was the best national and global economy in human history in pre-COVID 2019, with the most jobs at the highest wages ever.
He also mentions the push for “Universal Basic Income” as a response to AI.
Now we’re heading into the third such panic of the new century with AI, coupled with a continuous drumbeat of demand for Communist-inspired Universal Basic Income.
“This time is different; AI is different,” they say, but is it?
His take is that for many important parts of our economy, technological innovation is virtually forbidden now.
The red sectors are heavily regulated by the government and are steadily pushing prices up over time.
The blue sectors are where tech is able to push down prices and improve quality.
While AI will have a massive impact in some sectors, overall, a lot of our industry is heavily regulated and protected from technological advance.
(and that’s growing as a % of GDP)
Brave Browser Launches an AI Summarization Feature
If you haven’t heard of the Brave Browser, it’s worth giving it a look.
It’s a free and open source browser (competitor to Chrome, Explorer, Firefox and Safari).
It acts as an automatic ad blocker along with a host of privacy focused features.
It’s MUCH faster to open most pages in Brave than in Chrome, because Brave will by default not load a lot of the ads, scripts and tracking cookies that Chrome will.
Recently it launched a “AI Summarization Feature” where each search result comes with a “ChatGPT-like” result that briefly summarizes the result.
You can try it here:
Brave, while still tiny compared to Chrome and the other browsers, is the fastest growing search engine since Bing.
(Brave search and Brave Browser can be used independently, but are growing at similar rates)
Brave also has an interesting proposal for how to improve online advertising that includes using the blockchain to reward the users for browsing the web.
Users would earn a BAT (basic attention token) that advertisers would also need to buy to be able to advertise to users.
This would take control from the Big Tech platforms and advertisers and give it to the users (that’s us).
Brave is founded by Brendan Eich, the guy who co-founded Mozilla, so this project has a lot of money, brainpower and name recognition behind it.
This might be something to keep an eye out on, especially if the user growth continues and Google search starts getting slowly replaced with AI powered search.
The BAT token, is a crypto token (and I don’t recommend people gamble with Crypto), but it has a legitimate use because advertisers would need to purchase it to be able to run ads.
Already Brave has bagged advertisers like Verizon, Purple Mattress and Ashley Homestore.
Again, something to keep an eye on if it keeps growing.
LINK FOR MORE INFO:
https://brave.com/ – you can download the brave browser for free.
https://basicattentiontoken.org/ – this is the reward for browsing the web using Brave. Also the token that advertisers would need to use to purchase ads.
200-Year-Old Math Opens Up AI’s Mysterious Black Box
Whether it’s designing microchips or dreaming up new proteins, sometimes it seems like neural networks can do anything.
These brain-inspired artificial intelligence(AI) systems work in mysterious ways, raising concerns that what they are doing might not make any sense.
- A new study has shown that a 200-year-old math technique called Fourier analysis could help explain how neural networks solve complex problems such as predicting climate or modeling turbulence, leading to better accuracy and faster learning.
- Fourier analysis identifies regular patterns in data across space and time and the researchers applied it to a deep neural network trained to analyze complex turbulence in air and water flows.
- The combination of filters had been previously attempted to analyze climate and turbulence but with limited success until neural networks learned how to use them effectively.
- The researchers have developed a general framework to apply this approach to any physical system and any neural network architecture.
- Transfer learning, which re-trains key neurons to help neural networks analyze other systems, may benefit from this approach as the study suggests that retraining the shallowest layers close to the input may result in better performance.
- Understanding how neural networks learn to combine filters to reach the best results is the next step for the researchers.
The inside story of how ChatGPT was built from the people who made it
The team behind OpenAI’s ChatGPT speaks about the development, launch, and reception of the chatbot.
Initially released as a “research preview” in late 2022, ChatGPT quickly became a viral hit.
OpenAI has since updated the chatbot multiple times, implementing adversarial training to prevent users from manipulating it, and they are continuing to address issues such as bias and “jailbreaking.”
The success of ChatGPT has sparked a gold rush around large language models, with investors and companies worldwide jumping into the action.
READ THE FULL STORY HERE (ARCHIVED VERSION)
Is OpenAI making money from ChatGPT API?
OpenAI reduced their cost for users of ChatGPT by almost 10x… and then AGAIN by almost 10x in the span of a year.
Can they possible be making a profit on this?
This back of the napkin math says “yes!”.
Looks like the cost still makes this a profitable venture (many responses have disagreed, but overall it seems that this wasn’t a “cash burn” by OpenAI).
Stocks like Google, Microsoft, Amazon and Nvidia are likely to see an impact from OpenAI.
The chatbot’s capabilities have led to speculation that it could become a threat to Google’s search engine, as it generates accurate search results and can provide in-depth analyses and suggestions.
Microsoft’s continued investment in core innovations such as its investment in OpenAI to bolster its search, cloud, and productivity software capabilities, as well as its deployment of capital towards expanding Azure, reinforces the sustainability of its growth trajectory.
Google’s LaMDA language model remains in the research phase and features only 137 billion parameters, compared to GPT-3’s 175 billion parameters.
Nvidia could benefit from the growing demands from the high-performance computing (HPC) and hyperscale data center segments, as its GPUs are in high demand.