andy99

46 ms·

961.
Uber offers 20 hours of free childcare in bid to lure female drivers(bbc.com)
2 points by andy99 2y ago|0 comments
962.
by andy99 2y ago
The problem is the overloading of the word "safety", not really the approach. If there was some mortal danger stemming from this yeah, maybe more objective oversite is needed. But it's a stunt anyway and there is no danger in
963.
by andy99 2y ago
All current "guardrails" are silly censorship / political correctness stuff, or for business appropriateness. They are also trivially circumvented. There is no "threat" from the un-shielded capability of current or
964.
by andy99 2y ago
Existing laws cover almost everything "bad" you could do with AI/ML. It's not like there's some "I used AI" loophole that exempts one from the law. So most of this is about either regulatory capture, self
965.
Vancouver council to consider supporting 'pod hotels' in vacant office space(bc.ctvnews.ca)
1 points by andy99 2y ago|0 comments
966.
by andy99 2y ago
I use iterm2 as a default and haven't given it any thought or looked into to it. But "AI powered" very often implies spying on you and sending home everything it can which has me a bit concerned. Is this true or likely to bec
967.
by andy99 2y ago
100%, I'm glad this is starting to get some mainstream attention. Unfortunately I think "open source" is a lost brand, there's already been too many shenanigans and even on the "good" side of the debate there&#
968.
by andy99 2y ago
I'd be interested in seeing the marketing intersection between "AI powered" and "AI free". Lots of room to fudge either way.
969.
by andy99 2y ago
That's clearly a typo, not a confabulation in the usual sense. You could really even frame how it happens the same way. There's some small, nonzero probability I'll hit an adjacent key and make a typo that's context depe
970.
by andy99 2y ago
We launched the first Voyager in the 70s. 50 years on do we have anything catching up to it?
971.
by andy99 2y ago
It's happened with academic journals like Science and Nature too...
972.
by andy99 2y ago
If they clarified with examples people would laugh at it and not take it seriously[0]. Better to couch it in vague terms like harms and safety and let people imagine what they want. There are no serious examples of AI giving "dangerous
973.
by andy99 2y ago
In fairness, workforce strategy changes should be decoupled from financial performance except in extreme cases. Making money is not a reason to keep people you don't need, nor is losing money a reason to cut people as long as you can
974.
by andy99 2y ago
I feel like there should be something in Montreal that would fit in here but I can't come up with a good example. The habitat '67 is close but doesn't have enough vegetation. There's various decaying concrete stuff in th
975.
by andy99 2y ago
And if you want to understand I'd recommend this post (gpt2 in 60 lines of numpy) and the post on attention it links to. The concepts are mostly identical to llama, just with a few minor architectural tweaks. https://jaykmod
976.
by andy99 2y ago
https://archive.md/6GyYi
977.
by andy99 2y ago
> the censorship necessary for corporate applications can be trivially implemented as a toggleable layer, using a small, efficient, specialist model to detect no-no words and wrongthink in inputs/outputs. What's really funny is
978.
by andy99 2y ago
Just to add, they don't have the baggage of google or Meta so they can do more without worrying how it impacts the rest of the company. And of the big players they seem the most aware of how important good data is and have paid for l
979.
by andy99 2y ago
I don't want to be dismissive, it's a fun project, but this has been done a lot already - maybe not with llama3 but the architecture is basically the same as llama2. Look at the big list of from scratch implementations on Karpathy
980.
by andy99 2y ago
Other than the appealing symmetry of being able to use a language to parse that same language, this feels very similar to https://stackoverflow.com/questions/1732348/regex-match-open... Edit: I just checked and sa
981.
by andy99 2y ago
"How do you know we're not just next token predictors" is the thought terminating cliche . We know that's what LLMs are. It was certainly eye opening to see how far that gets you. But any deeper claims about intelligence
982.
by andy99 2y ago
Luckily that doesn't exist. It's about equivalent to spinning up a Ghostbusters unit and letting it wither. You can argue that we'll be in trouble if we encounter ghosts, but on likelihood weighted basis there are much better
983.
by andy99 2y ago
Groupthink and heavy pro-establishment bias
984.
Colorado Enacts Law Governing Artificial Intelligence Systems(fpf.org)
2 points by andy99 2y ago|0 comments
985.
by andy99 2y ago
They'll realize that weakness eventually and replace them with spider like robots that are more mobile and harder to take down.
986.
by andy99 2y ago
Going to turn out like every other no code / low code tools, works great for a demo of something you'd never do, customizing to a real application is just as much work as coding it
987.
by andy99 2y ago
An Arizona woman has been accused of helping generate millions of dollars for North Korea’s ballistic missile program Obviously the whole thing is illegal regardless but there was no evidence of the above claim in the article. I can
988.
by andy99 2y ago
In practice it doesn't work that way. What's downsized is line workers that exec's treat as fungible and don't care about anyway. Any "savings" goes to more management to run whatever the replacement solution i
989.
by andy99 2y ago
No. People that boil a job down to a simple cognitive task rarely understand what the actual work of the job is.
990.
by andy99 2y ago
I probably missed the point, but open source is a license granted by a copyright holder. It's orthogonal to any project structure. If there's a collaboration of any kind, it would take place under a separate agreement, it's n

More ›