Viewing profile — andrewdb

Why do we call them GPUs these days? Most GPUs, sitting in racks in datacenters, aren't "processing graphics" anyhow.

Sure, there is a problem with slop AI PRs _now_ . That will not remain true for infinity. What happens when the AI PRs aren't slop?

We are getting to a point where AI will be able to construct sound arguments in prose. They will make logical sense. Dismissing them only because of their origin is fallacious thin…

If the PR had been proposed by a human, but it was 100% identical to the output generated by the bot, would it have been accepted?

So when will multiple Waymo cars communicate input data to one another to avoid the blind spots? This would give the ability to see things other cars cannot see as well.

One way to slightly mitigate the difficulties of nuance in language when translating to formal arguments is to attwmpt to always steelman the argument. Afford it all the guarded la…

A prompt that I like to use for this: --- Intake the following block of text and then formulate it as a steelmanned deductive argument. Use the format of premises and conclusion. A…

Not just chip exports. It also limits model weights. From the article: In addition to the semiconductor controls, the new rules also limit the export of closed AI model weights, wh…

In large part, yes

Steel-manned Deductive Argument Premise 1: AI (specifically, statistical modeling based on hidden layer neural networks) has been increasingly integrated into various technological…

Those are good suggestions. I will use some of them! It is also interesting to go back and forth with the model, asking it to mitigate fallacies listed, and then re-check for falla…

I have found the below to be a good starting point for formulating text into classical formulated arguments. Intake the following block of text and then formulate it as a steelmann…

If the nutritional value of a plant's output is governed by its genetics, then this should be a solvable problem. Using the boosted method, breed a high nutritional value plant wit…

If only this Dockerfile were real. It would greatly help app developers and publishers: FROM apple/mac-os-slim:latest ... Please, Apple, please let your developers use more virtual…

An over-simplified approach that would be interesting to explore: 1. Given a collection of logical arguments, figure out how to assign a hash to each argument. 2. Build a Merkle tr…

Steel-manned deductive argument Premise 1: The consumption of energy and natural resources required for AI infrastructure, particularly if it continues to grow, is massive and pote…

Possible fallacies in the post's argument Slippery Slope: The argument may overstate the direct line from investment in AI to catastrophic outcomes like global financial depression…

Grats on the launch! Would be very interested to see this working with GCP Cloud Run.

A good test of LLMs: > Give me a list of 5 words having 5 syllables each Mistral 8x7B and GPT 4 both choke on this frequently.

"I don't play craps. I'm a dice engineer!" - prompt engineers, satirically

yes.

I assume you're talking about mDAU. This is what Twitter says, essentially: "We used internal data to produce a judgement-driven estimate that something is X, but we could be wrong…

Not only is the trial schedule not changed, but Chancellor McCormick also made it clear she had little patience for expansive discovery for the whistleblower points. She is only al…

Musk has the burden of proof to show fraud or MAE (material adverse effect). For fraud, he needs to prove scienter as well as reliance. High bar, and undermined by his own tweets a…