bart_spoon

10 ms·

61.
by bart_spoon 3y ago
Seems like that only applies if you anthropomorphize the AI and consider it the author, rather than a tool utilized by the artist. I mean, yes, the AI is doing the bulk of the work, but so is photoshop for a lot of digital art.
62.
by bart_spoon 3y ago
Ok but who is actually saying the content/product result isn’t important? Seems like a strawman.
63.
by bart_spoon 3y ago
And yet isn’t Pixar famous for their intensive, structured process for movie development?
64.
by bart_spoon 3y ago
The opposite is also concerning. IP law has always been convoluted, messy, contradictory, and morally ambiguous. The complaints of IP violation by LLMs are simply taking these inherent flaws and making them immediately obvious, forcing deci
65.
by bart_spoon 3y ago
The concern many MtG players have is that throwing a bunch of other IP crossovers into the card game has a variety of potentially negative effects: - established players who enjoyed the theme of Magic in the first place might be turned off
66.
by bart_spoon 3y ago
There’s only two camps, investors and gamers. I think that anyone who wants to “have cards that have relative value” are investors, whether they realize it or not. Its one thing to collect because you like something inherent about the cards
67.
by bart_spoon 3y ago
“Cure” is a tough bar, but I believe Paxlovid, the anti-viral used to reduce Covid severity, was identified using ML. There’s many companies like Recursion Pharma which are entirely focused on using ML for drug discovery, and from what I ca
68.
by bart_spoon 3y ago
It’s the cost of compute hardware required to train a model of that size
69.
by bart_spoon 3y ago
Understanding meaning what?
70.
by bart_spoon 3y ago
DBT is open source and can be used entirely for free. They sell a hosted version. Also DBT and Airflow are different tools, aimed at solving completely different problems. The Airflow competitor in this article is Dagster, which is also ope
71.
by bart_spoon 3y ago
Far more helpful to me than these essays on the benefits of a particular paradigm is a simple, minimal example in a repo that I can dig into and explore.
72.
by bart_spoon 3y ago
Reading through this entire exchange and it would seem your argument inevitably falls back on this argument of consciousness, which is both arbitrary (why does that matter in the original context of the distinction in learning, particularly
73.
by bart_spoon 3y ago
Feel free to expound on the important differences
74.
by bart_spoon 3y ago
> For example any discussion about copyrighted works, which is a hot topic, will inevitably end up with someone equating an LLM "learning" to a human learning, as if the two are identical. I don’t think the point is that the tw
75.
by bart_spoon 3y ago
That is an oft repeated myth. A recent study by UCSF of California homelessness found that 90% of homeless Californians became homeless while already living in California, and 75% still live in the same county as they did when they became h
76.
by bart_spoon 3y ago
Altman’s departure represented a real possibility they may miss out of enormous amounts of wealth. People are willing to put up with or even support a lot of things when money is on the line.
77.
by bart_spoon 3y ago
I think a non-LLM model would be good for learning to draft within a particular set, because it can simply encode each card and learn what cards do well together, comparative strength, etc. without learning the stats/rules text on th
78.
by bart_spoon 3y ago
This has been true for a lot of publications I used to like to read. Wired, Time, Sports Illustrated all are much, much worse today than I remember them being 20 years ago. I don’t know if that’s me or a genuine change in their quality.
79.
by bart_spoon 3y ago
One doesn’t need arms and legs or any physical presence at all to do immense damage to society. Ransomware attacks cripple companies, hospitals, utilities, and governments on a regular basis. The US and Israel have demonstrated, in at least
80.
by bart_spoon 3y ago
So according to you, human beings are considered to possess intelligence prior to grade school?
81.
by bart_spoon 3y ago
Many children (and adults for that matter) would fail to qualify according to that definition of intelligence.
82.
by bart_spoon 3y ago
Transformers came about in 2017, so extend the time horizon by a whopping 2 years and OP’s point remains. AlexNet was 2012, which is what essentially kicked off the deep learning neural network revolution. The original implementation of Alp
83.
by bart_spoon 3y ago
The conflation of concern over AI safety and effective altruism is bizarre and unwarranted in my mind. Just because there is overlap between adherents of the two does not make them the same. Effective altruism largely seems like billionaire
84.
by bart_spoon 3y ago
Isn’t that reference like the exact opposite? It’s nominally self-aggrandizing (he’s the only one who could have gotten it right) but is actually about a higher goal (sacrificing himself for everyone else’s sake). That’s how I remember reac
85.
by bart_spoon 3y ago
Stock options?
86.
by bart_spoon 3y ago
Microsoft has already done major layoffs over the last year of their own employees. Why wouldn’t they lay off OpenAI employees?
87.
by bart_spoon 3y ago
That is entirely dependent on how that wealth is obtained
88.
by bart_spoon 3y ago
Perhaps because, for all of Silicon Valley and the tech industries platitudes about wanting to make the world a better place, 90% of them are solely interested in the fastest path to wealth.
89.
by bart_spoon 3y ago
Was it due to incompetence though? The way it has played out has made me feel it was always doomed. It is apparent that those concerned with AI safety were gravely concerned with the direction the company was taking, and were losing power r
90.
by bart_spoon 3y ago
> The exchange of services and goods in a market is positive-sum. As long as you ignore externalities, yes.

More ›