frabcus

11 ms·

151.
by frabcus 2y ago
Who on social media, and paid to do investigations how? And with whose algorithm recommending it with what incentives? It's not journalism to only read the primary sources you have time to read. That's not proper research, it&#x27
152.
by frabcus 2y ago
I didn't know before reading the Wikipedia article you linked that Malthus's essay led directly to the creation of the UK census. Which is one of the mechanisms by which we cope with population growth! Perhaps both things can be t
153.
by frabcus 2y ago
Me too! But I can't really remember why. The proof was impressive.
154.
by frabcus 2y ago
I think if models improve (but we don't get a full singularity) then jobs will increase. e.g. if software is 5x less cost to make, demand will go up more than 5x as supply is highly limited now. Lots of companies want better software b
155.
by frabcus 2y ago
I suspect people's internal mind abilities impact this a lot. I have little visual memory so icons are hard for me. Others I think struggle with quickly scanning words. I think both are needed as we are varied people, and empathy for t
156.
by frabcus 2y ago
No, it's not, as it means nobody is pushing for actually open models. A truly open model has open code that gathers pre-training data, open pre-training data, open RLHF data, open RLAIF data generated from its open constitution and so
157.
by frabcus 2y ago
This is tricky as I think it is uncertain. Right now the answer is user experience, customs workflows layered on top of the models and onboarding specific enterprises to use it. If suddenly agentic stuff works really well... Then that break
158.
by frabcus 2y ago
The point presumably is that the valuation in a round is paper money. Whereas Elon's group have just offered cash, so they've set the actual value. The law might only care that the non-profit gets a market cash value actually bid,
159.
by frabcus 2y ago
Anthropic was the only company in that list not to have paid for their CEO or founder to attend the inaugeration of the current ruler of the exectuive branch of the US a week ago.
160.
by frabcus 2y ago
I don't think it needs to be deterministic - and if it isn't, having the data and code becomes even more important! Compilers generally aren't deterministic (see the reproduceable build movement), yet we still use their out
161.
by frabcus 2y ago
Pyright with Python is lovely too!
162.
by frabcus 2y ago
Python is slightly above it on the StackOverflow 2024 developer survey. https://survey.stackoverflow.co/2024/technology#2-programmin...
163.
by frabcus 2y ago
There are tasks in Traitors! They're done in a group with some things that traitors have different incentives for. But yes it makes it more like Among Us than werewolf!
164.
by frabcus 2y ago
I remember the version of this called Pangolins on the ZX Spectrum. Because pangolins were an unusual case (scaly mammals). Looking it up, apparently it was a type-in in the programmling book that came with the computer. https://
165.
by frabcus 2y ago
One of those social networks has a protocol and lets end users make their own feed algorithm and moderation system. The other never has. There can be technical differences between networks as well as social.
166.
by frabcus 2y ago
This one is at least a Benefit Corp. It's worth a go - seems to be working for, say, Kickstarter, which hasn't enshittified yet.
167.
by frabcus 2y ago
Retweet was original using the text "RT". Hashtags were also invented by Twitter users.
168.
by frabcus 2y ago
It means I have to remember to put it on charge before I leave the office, and I never do remember. It stresses me out even trying to remember. My current strategy is to set a calendar alarm at 6pm once a month to remind me to plug it in as
169.
by frabcus 2y ago
It could poison the well - make it hard for people to find new good podcasts, and reduce discovery and revenue. Also they could fragment our society even more, disconnect people from people. Doesn't seem worth the risk. If people want
170.
by frabcus 2y ago
Personally, I hate even the idea of an AI made podcast, because to me podcasts are personal and emotional. They're about the individual humans who make them. They're not just a source of "information".
171.
by frabcus 2y ago
Hmmm - it's been well known and talked about when I've worked for large B2C apps! In London.
172.
by frabcus 2y ago
The answer a few years ago was nothing (eg just use Railway or other actual PaaS) or Terraform. Anything else mature and worth trying? Although I guess you mean individual machine config not operations in general!
173.
by frabcus 2y ago
Must include the code that curates the data
174.
by frabcus 2y ago
Meta could change the license of future releases of Llama and kill your business built on it. If the training data was openly available, even if you can't afford to res train a new version, a competitor like Amazon could do it for you
175.
by frabcus 2y ago
Reproducible builds of software binaries are a thing, but they aren't routinely done. Likewise training an AI is deterministic if you do it the same each time. And slight variances lead to similar capability models.
176.
by frabcus 2y ago
They could release the code that gathers and curates the data. Give a reproducible system for getting the pre training data. And presumably they own the post training RLHF stuff so could open that. Without those you're locked in to the
177.
by frabcus 2y ago
It is open weights not open source. If you can't train it and don't know the training data and can't use it to train your own models, it is a closed model aa a whole. Even if you have the binary weights.
178.
by frabcus 2y ago
If it was actually open source with data and the data curation code releases, they wouldn't be able to abuse it the same way. It is open weights, closed training data.
179.
by frabcus 2y ago
Also it isn't source code, it is a binary. You need at least the data curation code and preferably the data itself for it to be actually source code in the practical sense that anyone can remake the build. Llama could change the licens
180.
by frabcus 2y ago
Unfortunately it is barely more open source than Windows. Llama 3 weights are binary code and while the license is pretty good it isn't open source.

More ›