Weekend Reads: Five +1 Links
GenAI is OUT. Agents and workflows. Aspirations of the tech-gods. Jean-Luc Picard.
Generative AI is OUT. Reasoning is IN.
First, you should follow The Algorithmic Bridge. The author, Alberto Romero, consistently brings fresh updates in a very consumable way - he posts a lot too.
This article is talking about the updates to OpenAI's new o1 series and how it capabilities shift the conversation from GenAI to enhanced reasoning models. TL;DR
Chat bots are cooked
The divide between AI-Rich and AI-Poors is getting wider
Human inferiority complex
Communicating the Value Prop
OpenAI o1: 10 Implications For the Future
5 Levels of AI Agents
I found Cobus Greyling's description of the characteristics of AI Agents to be very insightful. He outlines 22 similarities and differences between Agents and Workflow automation.
AI Agents v. Workflow automation
How AI Could Transform the World for the Better
Great read and the latest from Dario Amodei, CEO of Anthropic.
Largely, his predictions on where AI could take us are inspiring and aspirational. I say Inspiring in the sense that it is the most human (user) focused I have read to date.
The potential AI impacts he outlines range from Biology and physical health to work and meaning.
He maintains that in the short term, humans will remain relevant in economic value but long term, our economic structure may need to be reevaluated.
Amodei envisions that as AI becomes more capable and cost effective, humanity will shift towards a future where human cooperation and compassion can thrive.
Hallelujah and amen to "Machines of Loving Grace"
Dario's vision vs. Others
I found this article to be a good "devil’s advocate" read to Dario's vision. The responsibility CEOs of place upon their own shoulders is massive. But who put them in charge?
AI companies, led by figures like Sam Altman of OpenAI, are pursuing the development of artificial general intelligence (AGI).
This is raises obvious concerns about the lack of public consent and democratic oversight. Critics are arguing that the rapid advancement of AI and the trajectory towards AGI is occurring without adequate societal input, which could lead to significant existential risks.
The article discusses objections to the notion of "permissionless invention," emphasizing the need for and avenue of informed consent and public engagement in decisions that could fundamentally alter society.
I like where tech is leading us and it's healthy to look at who is leading tech.
AI companies are trying to build god. Shouldn't they get our permission first?
A playground for accelerating imagination
Superstudio by Kaiber is a creative platform designed to help imagination during all phases of exploration by allowing users to generate and manipulate images and videos through various "Core Flows".
Users can create Core Flows by mixing Elements, curate their creations into Collections, and explore their ideas in depth.
For around $15 a month, Superstudio offers a pretty comprehensive toolkit for turning creative visions into more creative visions.
And I really like their patch and node and infinite canvas interface.
What character did you get?
Reddit is a great source of information. You have to sort through a lot of brain rot, but I really liked this prompt I found in a post:
"Based on our interactions so far, what fictional character do I share the most personality traits with?"
I got Jean-Luc Picard from TNG. If you've never watched this, drop everything and watch. You have all weekend to get through 7 seasons (and 4 feature films).
What character did you get?