10xDev

21 ms·

301.
by 10xDev 4y ago
Good thing we got some of our smartest people working on generative models, the outputs of which are being banned across the web instead of solving intelligence as was the original goal.
302.
by 10xDev 4y ago
It's much less the name and much more the recent hype around chatGPT. And no one can deliver to the ridiculous levels of hype 'influencers' are flooding on social media for engagement.
303.
by 10xDev 4y ago
Building a PC doesn't make much sense anymore for most people now with cloud computing, Steam Deck, Apple silicon and GPU prices being ridiculously high.
304.
by 10xDev 4y ago
Just get a steam deck and use EmuDeck.
305.
by 10xDev 4y ago
Is the data incorrect or somehow misleading? If the answer is no to both of these then your point is irrelevant.
306.
by 10xDev 4y ago
Sure, when you are a supplier of tools this is a great example. However, to point this as a shining example for all eCommerce sites is simply misunderstanding the difference in target audiences and products. For example, a person may visit
307.
by 10xDev 4y ago
I find it exciting but the overload of cherrypicked examples with the exaggerated context that surrounds it is certainly exhausting.
308.
by 10xDev 4y ago
Yeah I'm sure they didn't think of that themselves. Posts like this are to generate discussion not to find a solution.
309.
by 10xDev 4y ago
A key component to intelligence is generalisation which ironically you failed to do with the examples I gave.
310.
by 10xDev 4y ago
Let me know when it scales.
311.
by 10xDev 4y ago
It is not as constrained but still fairly constrained and with a lot of data. And I was referring to the comparison to children with a search engine.
312.
by 10xDev 4y ago
No, again we just discovered it only does well in constrained environments with well defined rules and a lot of data. Either that or something subjective that it can afford a lot of inaccuracy without looking totally stupid. It literally ca
313.
by 10xDev 4y ago
No in Tesla's case it is because a person still has to be on the wheel to correct any potential errors.
314.
by 10xDev 4y ago
People seem to be especially impressed by code generation like people were with chess because we associate those tasks with intelligence but they are constrained environments with well defined rules and a lot of data. Anyone who has followe
315.
by 10xDev 4y ago
No matter how much the speed of development increases, the problem becomes significantly harder with a limit to how much transformer based models can do. There is a reason why you aren't seeing GPT-4 despite the rapid progress we saw w
316.
by 10xDev 4y ago
chatGPT spits out the first thing to be generated with no self-awareness that it could be wrong let alone have self-reflection to correct itself. Maybe it seems people can be that way but this is generally not how humans function.
317.
by 10xDev 4y ago
I am not claiming anything other than the fact that we do not know the training data therefore not much can be inferred about how well it generalises from some success case.
318.
by 10xDev 4y ago
We don't know how well it is generalising to your input without knowing the training data. Failure cases however tell us a lot, such as the following: Input: "If five cats can catch five mice in five minutes, how long will it take
319.
by 10xDev 4y ago
There is work on zero-shot learning and on procedurally generated environments like Procgen but it doesn't go mainstream because the results aren't as cool to most as some new record on a known board game or video game.
320.
by 10xDev 4y ago
That's a fancy wikipedia chatbot. Nothing to do with generalisation.
321.
by 10xDev 4y ago
Chess variants: Chess but worse.
322.
by 10xDev 4y ago
> you win by forcing your opposition to mess up You just described all of chess.