jvalencia

7 ms·

31.
by jvalencia 2y ago
Changing license terms, aggressive changes to the API to disallow competition, horrendous user experience that requires a support contract. I really don't think there's a limit to what I've seen other companies do. I general
32.
by jvalencia 2y ago
I don't trust an open source solution by a major player unless it's published with other major players. Otherwise, the perverse incentives are too great.
33.
by jvalencia 2y ago
And yet you're here?
34.
by jvalencia 2y ago
I think this is the wrong question. The question assumes that AI is a derivative to your code. The reality is that there will come a day, sooner than you like, when the AI can replicate the application you wrote without your code. In that w
35.
by jvalencia 2y ago
No Cassandra :-/
36.
by jvalencia 3y ago
The cost of training quickly outpaces the cost of development as context length increases. So hardware is cheap until it isn't anymore, by orders of magnitude.
37.
by jvalencia 3y ago
Yes, it might increase for those outside the city, but for those inside, it's overall downward pressure.
38.
by jvalencia 3y ago
For those who didn't make it to the end, it's worth noting that there's a fair bit of tongue in cheek in there. > We believe MIT deserves further investigation, and, with the festive season upon us, we intend to follow up
39.
by jvalencia 3y ago
This is an interesting point. If I were conducting layoffs, the seemingly extra things like open source contributions, may be the first to go.
40.
by jvalencia 3y ago
How does it compare to: https://msgpack.org/index.html ?
41.
by jvalencia 3y ago
It seems like you should be able to extract statements of fact from competing sources, say Fox and CNN. The overlap of facts would likely be true enough. The distinct sets then become color as to why those facts were included/exclude
42.
by jvalencia 3y ago
Weather is a complex mix of many systems. The traditional approach is to understand all the systems and add them together. Since we don't understand them all fully, we get a lot of chaos. The ML algorithm doesn't care about the s
43.
by jvalencia 3y ago
I love the number 27 3^3 2+7 = 3+3+3
44.
by jvalencia 3y ago
If you drove 10-12k views, the algos on TikTok would notice and it probably boosts well beyond the initial bump.
45.
by jvalencia 3y ago
In 1999 I regularly paid $10 to fill my tank. It takes $50+ for the same tank size today. I ate a meal with a drink at Taco Bell for $2-3.
46.
by jvalencia 3y ago
100% this. ML has been very hyped for a while now and having it was seen as a badge for the company. To be fair, ML is not also something that was historically central to a degree, so many people wanting to get into AI, even good engineers
47.
by jvalencia 3y ago
I would argue that true maturity is when someone can take care of themselves, recognize their inputs with gratitude, and ask for help when needed with humility, not expecting that someone else would take care of them AND being grateful even
48.
by jvalencia 3y ago
I see people posting on Facebook offering services in local neighbor groups. My suggestion would be to find your city's / neighboring city's local Facebook groups and post there. In particular, lower income communities are
49.
by jvalencia 3y ago
You can't use it for generating training data.
50.
by jvalencia 3y ago
I see the same thing re: not seeing this with traditional restaurants. I'm always shocked that taking Chili's to go is cheaper than fast food.
51.
by jvalencia 3y ago
From Nvidias point of view, this might be an investment in their own business via network effects rather than a straight up investment in Hugging Face. I hope so, because when the VC money goes, the billing will go through the roof.
52.
by jvalencia 3y ago
I would disagree. The problem here is the same problem that plagues contractors working in multiple global time-zones. Lack of clear communication and guidelines creates a mess. If there is the right structures and algorithms, I do think
53.
by jvalencia 3y ago
The combination of an open chatbot that you can freely text with and a very limited one that only talks about the subject it wants to is pretty interesting. There's a little bit of the uncanny valley in there, but only if you realize
54.
by jvalencia 3y ago
Netflix 2 will be an AI generated studio that makes big budget films cheaply and owns all the rights. The code will largely be generated. They will not need actors/writers/tech people in the way that netflix does. So their margin
55.
by jvalencia 3y ago
Would that equivocation not be a positive thing? When we said global warming, it caused huge political nightmares when things were cold. By saying we study climate change, we depoliticize it. We state these are the facts about climate ch
56.
by jvalencia 3y ago
frame.work
57.
by jvalencia 3y ago
For advertisers it isn't. Signals for purchasing that could be had for cheap are extremely valuable. Especially if they're reliable. If Amazon could get me a qualified lead for the half the price of Google Ads, that would be a g
58.
by jvalencia 3y ago
this
59.
by jvalencia 4y ago
It won't stop it, but will allow enforcement agencies to enforce. Otherwise, they have no legal recourse to do so.
60.
Python Basics Onepager(github.com)
40 points by jvalencia 4y ago|11 comments

More ›