Stay calm and keep thinking for yourself
My favorite resources for building a curious and critical frame of mind.
The ChatGPT hype is still in full swing, or so it seems, with rumors of OpenAI monetizing ChatGPT, Microsoft announcing that it will include OpenAI’s technology in its Azure ecosystem and possibly even Office, and Google profiling itself as the keeper of ethical and moral standards in AI.
It can be quite a challenge to sift through all this, plus the ‘10000 ways to make your business more effective with ChatGPT’ and ‘Oh look what me ChatGPT just wrote all by meself!’ posts to find the information that’s really helpful: information that makes you look critically at everything that’s happening right now. Preferably without turning you into a total doomsayer that thinks all this should stop right now. Because obviously, this is not going to happen.
So this week, rather than taking you along in one of my conversations-as-analysis through ChatGPT, let me share some of the resources that I use to stay both curious and critical.
Books
My permanent companion and continuous reread for the last 2 months has been On Bullshit, together with its companion essay On Truth by philosopher Harry Frankfurt. If you haven’t done so yet, go read them. They’re short, captivating, and they will give you the words to voice that slightly eerie feeling of ‘something’s off here’.
Algorithms of oppression by Safiya Noble, Ph.D. — a disturbing insight on how biased algorithms create systemic racism and propagate white privilege. Although the main topic here is search engines in their traditional form, the patterns and examples hold up for generative conversational AI interfaces just as much. Or perhaps more, because bias will be even less visible in generative AI interfaces.
It’s probably no surprise that Gary Marcus & Ernest Davis’ Rebooting AI is in my list as well. I guess that many of you have already read this, and that most of the contents will already sound familiar to you. If you haven’t: it’s a great intro into what’s troubling about today’s AI hype, not just with large language models, but with current AI in general. Plus another way to approach AI, one that’s less harmful & more trustworthy.
People to follow
- Georg Huettenegger — one of those people who’s simply in the know when it comes to conversational AI. His LinkedIn feed is a great source for timely and well-researched information on anything conversational AI, with a focus on innovation, ethics and looking beyond the hype.
- Emily M. Bender and Timnit Gebru — touchstones in times when the AI hype is very real and very big. Co-authors of the paper On the Danger of Stochastic Parrots — can language models be too big?, they have been instrumental in making the larger public aware of the fundamental challenges of large language models.
- Dr Alan D Thompson — his website Lifearchitect.ai, his YouTube channel and his newsletter are a wonderful source of basically everything AI, with a special focus on AI and human intelligence.
- David Shapiro — his YouTube channel is a great resource to learn about the practical side of all things GPT, ChatGPT and LLMs in general, with a nice mix of insights, interviews and hands-on tutorials.
Publications that I often read
- Techcrunch.com
- MIT Technology Review
- VentureBeat
- AnalyticsIndiaMag.com
Places I go to learn
I strongly believe that the best way to understand technology is to actively work and build with it, even if it’s just at a very small scale.
- OpenAI’s playground and intro tutorials — a good place to start experimenting with LLM technology
- Cohere’s blog — consistently delivers high-quality learnings and insights in a very accessible way, even if you’re a non-techie like me
- Machine Learning Street Talk — a YouTube channel that really deep dives into the fundamentals of many ML topics. I go here to learn about computer science, AGI, philosophy and often simply taking in so many insights from so many great scientists.