Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
robbomacrae
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
10 ms
·
91.
▲
by
robbomacrae
2y ago
It's a theory on how to optimally rebalance your investment portfolio every day. Original paper by Thomas Cover: https://isl.stanford.edu/~cover/papers/paper93.pdf A good breakdown with links to code examples
92.
▲
by
robbomacrae
2y ago
Perhaps but with a modern/futuristic twist. The cells in our body have one by one been replaced with new ones and the old ones have died off. This is not the case for our "talking" neurons which is what I was trying to limi
93.
▲
by
robbomacrae
2y ago
That all makes sense. But let's run with this way beyond the foreseeable tech. What if you can replace each neuron in situ one by one up to X%. Then what if it was reversible (the neurons were initially just bypassed, commented out). S
94.
▲
by
robbomacrae
2y ago
Or turn it into something wider than HN.. a button that can be added to any page on the internet that becomes easier to recall "hey I wonder who else likes or wants to discuss this article on meet.x?
95.
▲
by
robbomacrae
2y ago
I’m curious to know of similar examples happening or references that back this up if you or anyone has them.
96.
▲
by
robbomacrae
2y ago
I work at SoundHound. If it didn't get it in two tries it's likely we didn't have that song in the database. Both Shazam and SH have knowledge gaps.
97.
▲
by
robbomacrae
2y ago
Code, model, data and under Apache 2.0. Impressive. Curious how this was allowed to be more open source compared to Llama's interesting new take on "open source". Are other projects restricted in some form due to technical&#x
98.
▲
by
robbomacrae
2y ago
If it really is as you say then that sounds like it won't hold up when challenged in court but IANAL...
99.
▲
by
robbomacrae
2y ago
Whilst I agree with the reservations of the other replies I think you were implying in the future and I'm sure the LLM's will be more trustworthy and up to the task at some point. What I would really like to see now is all the new
100.
▲
by
robbomacrae
2y ago
These things might be able to produce comparable output but that wasn't my point. I agree that if we are comparing ourselves over the text that gets written then LLM's can achieve super intelligence. And writing text can indeed be
101.
▲
by
robbomacrae
2y ago
It's a cute generalization but you do yourself a great disservice. It's somewhat difficult to argue given the medium we have here and it may be impossible to disprove but consider that in first 30 minutes of your post being highly
102.
▲
by
robbomacrae
2y ago
Interesting timing. Microsoft Aurora[0] was announced yesterday although that got very little attention here[1]. Too recent to be added to the article. [0]: https://www.microsoft.com/en-us/research/blog/introd
103.
▲
by
robbomacrae
2y ago
Thank you for this tip!
104.
▲
by
robbomacrae
2y ago
Gmail pushes chrome as a default browser repeatedly too. No way to turn that off. The system knows you don't want it (having already said no 15 times) its just hoping the 16th time you will accidentally click yes. I suspect as search e
105.
▲
by
robbomacrae
2y ago
That's precisely what they did with Doodle God to imitate Morgan Freeman [0] and how James Veich deep faked David Attenborough in his PLnaT eRth video [1]. [0]: https://www.mercurynews.com/2021/07/20/how-
106.
▲
by
robbomacrae
2y ago
I think OP is just pointing out that "acting in the best interests of humanity" is fairly ambiguous and leaves enough room for interpretation and spin to cover any number of sins.
107.
▲
by
robbomacrae
2y ago
I feel like the lowest hanging fruit right now lies with the UI and already established techniques for reducing latency and making the experience smoother. Just executing the client right would give someone a competitive advantage right now
108.
▲
Apple to Power AI Features with M2 Ultra Servers
(macrumors.com)
5 points
by
robbomacrae
2y ago
|
0 comments
109.
▲
by
robbomacrae
2y ago
Having worked at Apple I can assure you it's not just spin. It's nigh on impossible to get permission to even compare your data with another service inside of Apple and even if you do get permission the user ids and everything are
110.
▲
by
robbomacrae
2y ago
They already do this. It's called federated learning and its a way for them to use your data to help personalize the model for you and also (to a much lesser extent) the global model for everyone whilst still respecting your data priva
111.
▲
by
robbomacrae
2y ago
I don't think this is what you meant but it matches the spec: federated learning is being used by Apple to train models for various applications and some of that happens on device (iphones/ipads) with your personal data before its
112.
▲
by
robbomacrae
2y ago
I'm generally a proponent of AI and LLM but to me the decision was the right one. You are tasking people with implementing an idea generated by an algorithmic model with (I'm guessing) zero oversight that might have very little tr
113.
▲
by
robbomacrae
3y ago
I, at least, had sources. I tried looking evidence of this new fact about Yugoslavia you presented but could not find any. It was called the Kosovo war. Why can we not compare the two? They seem VERY comparable to me. But if Wikipedia and A
114.
▲
by
robbomacrae
3y ago
Yes its genocide. In 1999 Yugoslavia killed ~12 thousand Albanians and displaced ~85 thousand more. Bill Clintons secretary of defense had no problem calling that genocide: "The appalling accounts of mass killing in Kosovo and the pict
115.
▲
by
robbomacrae
3y ago
Its not even "lite" genocide... In 1999 Yugoslavia killed ~12 thousand Albanians and displaced ~85 thousand more. Bill Clintons secretary of defense had no problem calling that genocide: "The appalling accounts of mass killin
116.
▲
by
robbomacrae
3y ago
Difference is someone has to fire those weapons and that someone is responsible.
117.
▲
by
robbomacrae
3y ago
It's a good question. I also immediately found myself asking the same one of myself after posting that comment. I guess part of me just wants as many possible breakpoints along the process as possible. But also at least then you have s
118.
▲
by
robbomacrae
3y ago
Disturbing indeed. I've been worried a push back in AI is coming and this sort of story could be a tipping point and certainly would justify a period of reflection. And your probably right that the alternatives maybe worse, the folks b
119.
▲
by
robbomacrae
3y ago
> Specifically, we encourage steps like: * Phasing out voice based authentication as a security measure for accessing bank accounts and other sensitive information I've been trying to tell people this for years. My bank still tries
120.
▲
by
robbomacrae
3y ago
Not trying to hijack this. Great demo! But STT can be very much real-time now. Try SoundHound's transcription service available through the Houndify platform [0] (we really don't market this well enough). It's lightning fast
More ›