Closi

41 ms·

811.
by Closi 4y ago
Everyone has limited time, and if AI assistance can increase the speed you can develop & iterate the product to better match user needs that is how it can result in a better product. Equally if it can help devs launch a month earlier, t
812.
by Closi 4y ago
> Yes, calculators do that. I'm arguing that "AI" does not let programmers write better code faster. It lets them write worse code faster, or better code slower. The idea isn't to write better code faster, it's t
813.
by Closi 4y ago
I think you missed their point - calculators don't exist in accountancy firms to 'encourage accountants not to think', they exist because they dramatically speed up accountancy and make accountants more productive. Sure you
814.
by Closi 4y ago
It's a matter of rendering a bunch of transposed points as a path and then applying a fill. Not particularly different, other than that I assume OP's challenge was to see how tiny they could make this and that might expand the cod
815.
by Closi 4y ago
In ML python is effectively the interface rather than the bit that is doing the heavy lifting. The interface is designed to be easy to use (python) and the bit that is actually doing the work is designed to be heavily performant (which is C
816.
by Closi 4y ago
I’m not sure this is fair - a lot of the performance optimisations have come from applied mathematicians rather than programmers, and python is not generally considered to be the bottleneck (it is the interface rather than what is running t
817.
by Closi 4y ago
Countries would be insane to ban companies like OpenAI. What a huge foot-gun that would be! Imagine if a country banned computers in 1979 to protect jobs, other countries would have run circles around them and it would have had huge economi
818.
by Closi 4y ago
This assumes that a future version won't have super-human coding abilities which can write code vastly better than any human. The language it writes in might not look anything like our current languages, because it might be able to wri
819.
by Closi 4y ago
You couldn’t even install an app on the first iPhone! (It’s a bit chicken and egg anyway - better get some hardware out so developers can build for real users)
820.
by Closi 4y ago
Oh man, if I could do SketchUp in ‘real’ 3D that would be a dream. Half of the annoyances in working in 3D are doing some changes and then realising you got the perspective wrong. Quite happy to still work with a keyboard and mouse sitting
821.
by Closi 4y ago
I'm not convinced - at least on current hardware this seems well positioned for the cloud. My tiny Alexa puck isn't going to run a 180bn parameter LLM that runs best on 10 graphics cards any time soon, however it can already call
822.
by Closi 4y ago
This comment might end up being a bit like "640kb of RAM is more than anyone will ever need!" GPT3 already used almost 1/5th of that figure at 180bn parameters, and PaML uses 500bn.
823.
by Closi 4y ago
> Gamers aren’t going to want to buy this thing unless you can plug it into a PC and have it function as a regular VR headset and the Meta Quest crowd isn’t going to shell out 3 grand. Maybe Apple has just worked out that they can't
824.
by Closi 4y ago
Butterfly keyboards, touch bar and force-touch were features rather than products - most of these were actually iterated out of the product, although the touchbar is still in the latest MacBook Pro's. As a side point, 3D touch was
825.
by Closi 4y ago
What do we mean by cognition, how do we measure it, and what aspect of cognitive performance are 6 year olds better at though? Because GPT4 would destroy a 6 year old at a verbal reasoning, quantative reasoning or an abstract reasoning test
826.
by Closi 4y ago
> It might at least currently be on the cognitive level of an overconfident 6yo in many regards, but a 6yo with more knowledge than any single person alive. I'm not sure I agree entirely with this - It's very rare to see 6 year
827.
by Closi 4y ago
GPT basically reads the text you have input, and generates a set of 'likely' next words (technically 'tokens'). So for example, the input: Bears like to eat ________ GPT may effectively respond with Honey (33% likelihood
828.
by Closi 4y ago
I don't think it's a conspiracy theory - lots of companies are posting record profits right now as their prices have risen above the cost price inflation, so profiteering is a valid fourth class of price inflation.
829.
by Closi 4y ago
Worth noting that every other product MSCHF has made has actually been real though. No reason to believe this will be any different just because they are an art collective.
830.
by Closi 4y ago
Facilitating fraud != committing fraud Most banks and payment platforms will facilitate fraud to some extent, but it’s their duty to reduce this with compliance procedures.
831.
by Closi 4y ago
Worth noting that the title of this HN post is a little misleading - and Hindenburg actually have not accused Block of fraud (only of facilitating others to commit fraud through their platform).
832.
by Closi 4y ago
It’s legal but you must talk about facts and not make anything up, and you must act without insider information. This is why the report is surgical in its wording.
833.
by Closi 4y ago
Misleading headline - the article doesn't actually call Block Inc a fraud. It just claims that Block Inc are overvalued, that their metrics are misleading, and that their platform is facilitating fraud.
834.
by Closi 4y ago
Sure, but you will get better results at a smaller number of parameters from a specifically trained model right now if you are trying to train/host it yourself. Remember that GPT3 is 175 billion parameters so many times bigger than bot
835.
by Closi 4y ago
Amazing to reflect that this used to be 'state of the art' for ML only three years ago.
836.
by Closi 4y ago
Gradual incremental changes until there is suddenly an explosion of change and adoption, if the PC and Internet are anything to go by. In my view it will probably change the way we work to a large extent in 10 years, and tech work will prob
837.
by Closi 4y ago
It’s unlikely we are close to the peak IMO - the rate of improvement and discoveries appears to be increasing rather than decreasing, implying that we are still at the start of the curve. Progress doesn’t usually immediately stop, you would
838.
by Closi 4y ago
The assumption in your post is that the ‘bullshitting’ won’t/can’t be solved within a few iterations. Given the decrease of this already in GPT 4, it’s entirely possible that this will be solved soon. We are still right at the birth of
839.
by Closi 4y ago
It’s amazing how quick some people on HN are to dismiss how impressive it is for a generic language model to be able to play chess at any level - or indeed other feats of ChatGPT. It’s utterly staggering that it can work as well as it can o
840.
by Closi 4y ago
In what way is that even close to a Ponzi scheme?

More ›