I Am Not a Robot

The darkest side of the technological toast, with some philosophical butter.

The hottest Substack posts of I Am Not a Robot

And their main takeaways

Generating SQL with LLMs for fun and profit

69 HN points • 22 Jun 23

🕹 Technology AI SQL Programming Machine Learning

Language models can generate SQL queries, but can also create malicious queries if not careful.
Running infinite loops or allowing data exfiltration are risks with generated SQL queries.
Consider restricting permissions, making the database read-only, and avoiding prompt injection to reduce SQL injection risks with language models.

Why I am starting a hardcore tech company in my 50s

55 HN points • 26 Jul 23

💼 Business Technology Startups Entrepreneurship Venture Capital Consulting

Started a tech company in 50s due to interest in AI and experience in information retrieval and coding.
Focused on building stable and reliable AI systems, despite current trend of fast-paced development fueled by VC money.
Plans to start as a service provider in AI to build resilient systems and eventually create a product, prioritizing problem-solving and teamwork over money.

Should you use OpenAI's embeddings? Probably not, and here's why.

71 HN points • 30 Mar 23

🕹 Technology AI Embeddings Language Models Benchmarks

Consider using lighter embedding models before heavier ones.
If you are using a large model like Instructor XL, then consider trying OpenAI's embeddings for blind comparison.
Be cautious using OpenAI's embeddings due to internet dependency and potential future changes.

AGI Doom and the Drake equation

31 HN points • 07 Apr 23

🕹 Technology AI Ethics Philosophy Risk Assessment Machine Learning

The AGI doom scenario requires multiple rare conditions to happen simultaneously.
Current AI technology limitations make the worst case scenario less likely.
It's important to understand the technology and gaps to assess the likelihood of AGI doom.

The path to autonomous intelligent agents: what needs to happen.

2 HN points • 23 May 23

🕹 Technology AI Machine Learning Automation Security Reasoning

There is a big gap between AI hype and current technology.
Building autonomous agents involves overcoming obstacles like speed and cost of inference.
Improving agent quality in reasoning is a challenging but crucial aspect.

Get a weekly roundup of the best Substack posts, by hacker news affinity:

ChatGPT plugins: OpenAI's attempt to kill Google

1 HN point • 13 Apr 23

🕹 Technology AI Plugins Competition Innovation Marketplace

ChatGPT plugins are aimed at making it more useful like Google by connecting it with external resources.
With access to ChatGPT's plugin store, users can easily enhance their tasks such as shopping or gathering information.
OpenAI's advantage over Google in developing ChatGPT plugins lies in offering early adopters favorable deals and incentives.