Do you use a "cloud ML platform" — something like SageMaker or Vertex or H2O? What for?
@fchollet
-
Two Continua: Retrieval, Synthesis, and Program Execution
By
–
"There is a continuum between reasoning and retrieval" is half-true. There are two continua: one between pure retrieval and novel program synthesis (with many intermediate points that consist of subroutine retrieval + partial synthesis), and one between exact program execution
-
The Age of AI: When Algorithms Begin Designing Our World
By
–
Everything object, every process you interact with has been designed by a person (when it isn't the result of natural evolution). You'll know we've entered the age of AI when we'll start being surrounded with things designed by algorithms.
-
AI as Interface to Complex Information Space
By
–
The fundamental role of AI going forward is to serve as our interface to the an increasingly complex and information-intensive world. AI models are maps of information space, and we will use them as guides.
-
Deep Learning with Python: 2017 Predictions Materializing Eight Years Later
By
–
Deep learning with Python, 2017 (first edition). Paragraphs written in 2016, technically. Eight years later, much of this has come true, and the rest won't take much longer.
-
On the Logic of Running AI Products at Heavy Loss
By
–
One has to wonder what the logic of running a product like this at a heavy loss is (as opposed to pricing it higher). Normally you'd do it to undermine competition and secure a monopoly, but in this case they already have a monopoly by default since they were first to market and
-
High Query Load Per User Drives Model Profitability at Low Price Point
By
–
Unexpected — at that price point and for that type of model I would have assumed narrow profitability. I guess it's because the query load per user is very high.
-
How I Think About LLM Prompt Engineering
By
–
I've published a new post on my Substack, first one in a while. It's about LLMs. https://
fchollet.substack.com/p/how-i-think-
about-llm-prompt-engineering
… -
Human-level AI proves more challenging than previously anticipated
By
–
Human-level AI is harder than it seemed in February 2023.
-
Misuse and Misunderstanding of “Reasoning” in AI Context
By
–
In the context of AI, "reasoning" has got to be one of the most misused and misunderstood term ever, right up with "intelligence".