m00x

31 ms·

511.
H3 – Outperforming GPT-Neo-2.7B with only 2 attention layers(twitter.com)
10 points by m00x 4y ago|0 comments
512.
by m00x 4y ago
> This is a misunderstanding of what dictionaries are for. They’re not authoritative sources of what words mean, they’re a record of all the different ways a word has been used. Source?
513.
by m00x 4y ago
> One of the challenges faced by ChatGPT was in seeing it produce conversational-style output instead of simply the most common words that it saw. Do you have a source for that? I've implemented GPT from scratch and I don't see
514.
by m00x 4y ago
Yeah, that's a good point. The general public has now taken AGI as the term he recognizes as AI, and AI as what he calls ML. Language is fluid, and I don't think he'll win this battle. Even if the current ML models aren'
515.
by m00x 4y ago
It depends on how you define intelligence, but the most common definition of human intelligence is the ability to have abstract thought. Intelligence, reasoning and AI are all very loosely defined words, and unfortunately neuroscience hasn&
516.
by m00x 4y ago
> Reason is the capacity of consciously applying logic by drawing conclusions from new or existing information, with the aim of seeking the truth. ChatGPT doesn't care what word it puts in front of you, and if it's true or not.
517.
by m00x 4y ago
That's probably because the definition of `robot` is different in almost every dictionary. Cambridge defines it as: a machine controlled by a computer that is used to perform jobs automatically Others say it has to look human to be c
518.
by m00x 4y ago
I think you missed the point of the article. What he says is AI is more alike to what the public calls AGI.
519.
by m00x 4y ago
They don't have any reasoning capability. There is nowhere in their architecture that would give them the ability to reason. All they do is predict what would be the most likely word to go next in the sentence, considering the current
520.
by m00x 4y ago
He's arguing that AI should be closer to what the public calls AGI. The semantics are important here, since people think that LLMs like chatGPT have intellect, and some have even gone so far as to tie an IQ to it. The truth is that the
521.
by m00x 4y ago
It depends on the eng. If you're a senior eng with on-call responsibilities, you're getting full time prod access.
522.
by m00x 4y ago
I wonder why they didn't include everyone. Would be nice to know why.
523.
by m00x 4y ago
Amazon treats their engineers so poorly that they have to set up training pipelines directly into middle schools. Maybe they think they can make them drink the kool-aid before they grow their critical thinking skills.
524.
by m00x 4y ago
Customers got an email from Circle immediately after they were able to "resolve" the issue. The full post-mortem was released recently. We got the email on Jan 4 or Jan 5.
525.
by m00x 4y ago
The employee had malware on his computer, and the hackers might've used a VPN that's located in a "safe" zone. Security is pretty lax at most companies, and the more employees you have, the worse it is.
526.
by m00x 4y ago
Github actions are pretty amazing. It can run most CI workflows.
527.
by m00x 4y ago
I was SA'd by the VP of HR at a previous company. I asked around, and apparently there is no one that can report that to except for HR, so in several companies HR can do anything they want.
528.
by m00x 4y ago
You can only do so much as a manager/leader. Some people don't want to work and are incredibly lazy. You can lead a horse to water, but you can't make it drink. Also, it wasn't their first recourse. They fired her, but t
529.
by m00x 4y ago
Isn't that what the File System Access API is for? It allows you to open files on your own machine near-natively.
530.
by m00x 4y ago
"I used to do drugs. I still do, but I used to, too."
531.
by m00x 4y ago
Any reason why it's collecting dust? I have a few RISC-V boards and I haven't had any issues with them other than having to compile almost everything from source.
532.
by m00x 4y ago
A $2T loss in asset worth is much different than $2T of liquidity. People often confuse the two, but they're incredibly different.
533.
by m00x 4y ago
Unless I'm mistaken, chatGPT is trained on word embeddings, so it can't really make spelling mistakes, unless those misspelled words are included in the vocabulary.
534.
by m00x 4y ago
There are also 1 in 6B people with an IQ of 200 or above, and I imagine they're not prefacing every public forum interaction with: "I have an IQ of 200, here's my take on..."
535.
by m00x 4y ago
Since it doesn't have an age, it does not have an IQ.
536.
by m00x 4y ago
Keep in mind this isn't a real person with an IQ of 200, that's what the model learned about how people who claim to have an IQ of 200 on the internet. Usually people with a high IQ don't mention their IQ.
537.
by m00x 4y ago
I think wasm + webgpu will probably take over game development since it's incredibly portable from the start, and has near-native performance. Bevy is directly integrated with both, so I'm really hopeful of its future.
538.
by m00x 4y ago
It's usually the case when they abandon some high risk features, or they lay off low performers, and want to message that the company is doing well, just re-adjusting.
539.
by m00x 4y ago
That happened to me with Amazon and Microsoft. I asked Amazon to never email me again, and then I get a "got it, removing you from our list", and then the same recruiter emailed me saying I was a perfect fit.
540.
by m00x 4y ago
Wouldn't you just use an off-the-shelf rendering and physics engine instead of working on your own?

More ›