carom

My dream for a parsing library / language is that it would be able to read, manipulate, and then re-serialize the data. I'm sure there are a ton of edge cases there, but the round trip would be so useful for fuzzing and program analysis.

Just to note though, source copyright extends to its compiled form. There is probably an analogue there for model weights.

I was just going to fill it out how I would normally live Self Care > No Medication and immediately failed. That makes it feel like it has an agenda.

This article seems to have scoped AI as LLMs and totally missed the revolutionary application that is self driving cars. There will be a lot more applications outside of chat assistants.

The incentives are very poor in art, yet artists still create. I don't need people making video games for the incentives, I want the artists.

Interesting interpreting those as individualist. First can be read as a concern for family. Second is community and society. Third is also protection of community, you would be making a choice to intervene (an individual would leave). Fourth also is not the individual but again, family.

Catastrophic AI risk is such a larp. The systems are not sentient. The risk will always be around the human driving the LLM, not the LLM itself. We already have laws governing human behavior, company behavior. If an entity violates a law using an LLM, it has nothing to do with the LLM.

My battery was going out on my 12 and I got an SE. It's a good experience. If you can get a thumb print one, I personally like it a lot more than face ID.

This is the primary researcher behind the input hypothesis. [1]

1. https://en.m.wikipedia.org/wiki/Stephen_Krashen

>Will at centre of legal battle over Shakespeare’s home unearthed after 150 years

What a confusing title. I read it as the home being unearthed after 150 years and that there was a will involved in an active dispute over this newly unearthed home.

TSMC gets their machines from ASML who licenses their technology from the Department of Energy. The US will be OK.

I think you have perfectly described OP's vision. A unification of build systems and CI pipelines.

The top two winning teams of that xTech AI pitch competition were not even AI solutions. It just seemed like a vehicle for the Army to now be able to award those companies non competitive contracts.

Here is my PR, it aligns perfectly with the project goals. It contains a backdoor as binary blob that will be loaded dynamically upon execution. The models are nowhere near catching this and it would get merged. Even more simply, a subtle bug leading to a vulnerable release. They do not have logic enough to catch this stuff.

The other comment articulates the points much better than I would have, but I have a large number of (tested) food and environmental allergies. It is a very logical explanation that the adjuvants in vaccines would cause the body to also train an immune response to other things.

There is also a massive profit motive for pharma companies and many hospitals, when you couple that with the revolving door between industry and government, it seems like a situation ripe for corruption.

I don't see the harm in removing aluminum adjuvants from vaccines (we all buy aluminum free deodorant!). I don't see the harm in not vaccinating children for things they are unlikely to come into contact with (i.e. hepatitis B). In fact, I think it would be good to make the change and see what the health outcomes are over the next 30 years. That is how we will learn.

Ok, hold up. This study came up on Reddit a few weeks ago and my wife linked it to me. A lot of the comments were similar about how vaccine skeptics will never be convinced by it. So, being a vaccine skeptic, I went and read it.

>In this primary analysis, except for Asperger syndrome (hazard ratio, 1.13 [CI, 0.89 to 1.44]) and atypical autism (hazard ratio, 0.94 [CI, 0.79 to 1.12]), estimates for the individual outcomes were incompatible with any increased risk, with the upper bounds of the 95% CIs below 1.00. [1]

My understanding of this, and I am a software engineer so take it with a grain of salt, is that this study failed to disprove a link between aluminum in vaccines and aspergers! There is another section where it appears they played with the hyperparameters of their study and ended up with a lower hazard ratio for aspergers (I believe by extending the analysis window to 8 years of age, but it wasn't clear to me).

>Except for Asperger syndrome (hazard ratio, 1.02 [CI, 0.93 to 1.12]) and atypical autism (hazard ratio, 0.95 [CI, 0.88 to 1.03]), estimates for the individual neurodevelopmental outcomes assessed were incompatible with any increases in risk, with the upper bounds of the 95% CIs equal to or below 1.00.

That is to say, after reading the study, I am not convinced at all. I would like to see a longer analysis period (e.g. to 50 years of age) as many things go undiagnosed until later in life. From my reading though, this study failed to disprove a link despite what all the popsci headlines are saying.

1. https://www.acpjournals.org/doi/10.7326/ANNALS-25-00997

Edit: I know I am going to catch downvotes for this, but please go read the study and let me know where I am incorrect!

There are studies that going gluten free has positive health outcomes for schizophrenic patients. I could see keto working by the same mechanism.

That first point is not comparing students, it is saying that the H1B visas issued that year all have jobs lined up (which is a requirement of the visa). Those jobs are what the new graduates would normally be competing for.

I'm not sure the take away for the first point that user standards are rising is correct. Could that also be the number of people making games is increasing? I say this because more highly rated games and a trending down of the average (more slop) could explain that as well. I think the idea that standards are rising would hold constant the number of games.

The American DoD is a bit like this. It is much easier to get R&D contracts if you have PhD next to your name.

I strongly disagree with this ifs take. I want to validate data where it is used. I do not trust the caller (myself) to go read some comment about the assumptions on input data a function expects. I also don't want to duplicate that check in every caller.

There are typos and rough grammar in the first few paragraphs and I am actually very happy about that because I know I'm not reading LLM slop.

Ah, thank you. I had, admittedly, not read the article in OP. I thought this was very vaguely referencing another situation.

Any more details on this?

It is interesting because there are a large number of players who feel at home in the Call of Duty voice lobbies. Some people's banter and camaraderie is other people's toxicity.

What is the ETA on the kit shipping if I pre-order one?

How is the support for LFS? Also, what backend language? I have some Go code for implementing an LFS server and auth but did not want to build a full code forge. All of the major Git hosts have woefully bad LFS management (e.g. if you want to purge a file from history have to delete the whole repository).

>“I did [a noun]” is grammatically correct English object-verb-direct-object sentence structure.

I'm going to need some examples here because I'm filling in nouns to that structure and it does not sound correct. "I did [a] water." Huh?

I don't know about that. I would bet my computer uses far less energy than the average vehicle in someone's commute.

I have a Sceptre from Walmart. No internet connection. I run all services from an Apple TV. You can definitely buy a dumb TV.