You sour pusses are wrong. This is a smart move that amplifies a brilliant team from cursor with serious compute, raising the odds Elon can get to the frontier, which is worth so much these numbers will all look like a drop in the bucket.
You sour pusses are wrong. This is a smart move. Cursor has a brilliant, capable team with serious model chops who will be able to boost the odds of AGI success. They also come with a revenue generating machine.
Restaurants try to make food you will remember and want again. Authors try to write books you can't stop reading. It's silly to imagine that any type of media would do anything other than seek to gain your interest and attention. It's our job to have personal hygiene and to control our information diet. This postmodern social construction perspective that tries to blame everyone for our problems is a lame approach to the problem.
I see where you are coming from, but this is different because:
1. The restaurant isn't lacing your food with cocaine
2. The author is incentivizing reading, which is generally a good thing
You are right, it is our job to control our information diet but when the ability to control is diminished through the consumption, then we have a problem which doesn't fit into the model of "get your shit together".
There's also a social responsibility you inherit when you start selling products that have harmful side-effects. Auto manufacturers have to comply with emission testing, drug makers have to prove efficacy, air conditioning manufacturers have to adhere to air quality standards etc. What do social media companies have to do? Very little if anything, and that's the problem. We've yet to find a counter-balance that works and protects the consumer, so we're in this era where we're trying to find out what we can do, but the landscape is changing so fast that we're trying to hit a moving target.
> 1. The restaurant isn't lacing your food with cocaine
Cocaine creates a much larger chemical change in the brain and can kill you in high doses so it’s not really comparable.
> 2. The author is incentivizing reading, which is generally a good thing
Says who? Why is reading inherently a good thing?
And all of those things are regulated because they have what most consider to be objectively bad effects on the body (with metrics based on cellular harm).
If you genuinely want to understand what we're talking about here, I suggest you read up on how certain behavior discharge dopamine on cue, how that builds tolerance and addiction. Then perhaps you understand the difference between this subject and the examples you gave.
I agree. People are on social media more because other alternatives for entertainment have dwindled in affordability, quality, abundance, etc. Perhaps the problem that should be addressed are the lack of quality alternatives rather than the industry tactics of a particular form of entertainment.
This is a reductionist take on the problem. If I knowingly spike your drink with habit forming chemicals, and I don't tell you about it, am I in the wrong?
Of course they want to hide the data. The public freaks out with absurd claims about it being the fault of a chat bot when someone does something crazy. Humans need to remain 100% accountable for their own actions, and we should stop with this post-modern, social construction nonsense that pretends we are all like ping-pong balls just bouncing around between external forces.
I find it hard to believe that you fail to understand that vulnerable people can be influenced and manipulated into acting against their own welfare. Domestic sexual abuse of children is the simpler-to-understand form of this dynamic. Where emotional needs are exploited to direct the person's behavior to your personal benefit.
It is in these situations where the human performing the manipulation is the one responsible and not the victim.
Put another way, if you discovered that you were incidentally killing children each time you drove down a particular road, would you choose an alternative road or drive faster in an attempt to avoid detection?
Calling a chatbot yes-man a manipulator in this case is unreasonable. If you take a drama class and then believe you have become the King of England and try to attack Scotland, that isn't the fault of the drama teacher.
There are all kinds of products we know people will misuse for violence (guns, cars, knives), but we do not hold makers of these products accountable because it isn't reasonable to blame them for what a fringe minority do.
> Humans need to remain 100% accountable for their own actions
People undergoing psychotic delusions are definitionally not 100% culpable. If you say that people are 100% culpable despite mental state outside their control, I'd like to have you sign some things after you drink this scopolamine.
> it being the fault of a chat bot
It's contributory negligence. The chat bot could be designed to recognize psychotic delusions and urge the individual to seek help. Instead, it is negligently allows to reinforce those delusions.
The devs mentioned they're not currently looking into it. The game uses Vulkan which isn't supported by MacOS, so they'd have to write a whole second renderer just for Mac.
Bummer to read that MoltenVK is too buggy to use. ISTR that neither wgpu nor Dawn use it though in favor of their own WebGPU -> Metal backends, so maybe I shouldn’t be surprised.
Vulkan is a powerful API but it’s not universally cross platform like OpenGL.
You can use Whisky [0] which is a free tool that uses WINE to run Windows applications on macOS, including games. There are also paid tools like CrossOver and Parallels but Whisky works well enough for most use cases.
This is a great idea. Being able to dynamically scale up model sizes as datasets and use cases expand without needing to retrain from scratch could enable a Cambrian explosion of interesting stuff building on top of a Llama type model trained in this way.
It just makes sense to put the sensors near the phenomena being sensed, and if people know where they are they could be manipulated. If guns are going off and people are getting caught, that seems like effective policing, not over or under policing.
Seriously, it hurts. It makes sense, but it hurts. If only there were a more gentle path to editor modes. Maybe some simple graphical representation of the modes and commands that could be down in the corner? Like a dynamic vim infographic that clued a user into the most likely commands.
For me, the reason to learn vim is because if you know how to use it it's a really nice editor that's preinstalled on almost any system you shell into. I use VSCode on my home system and occasionally on a remote system, but I'll use neovim when doing a quick edit from the command line and vi when I happen to be o a remote system. In my ideal world, I would just use one editor everywhere, including if I'm hot-seating on a system that is not mine and without internet access: having vi already under my fingers is a nice approximation of great editor + everywhere.
If you learn Vim you’ll have access to Vi and Vi-like interfaces in lots of other software including terminals, database clients and everything that uses readline.
Curious what kind of interfaces you're thinking about? (Aside from vi, vim, nvim)
From my experience anything with "vi navigation" basically just means using the home row keys for navigation + modes. So I haven't come across many interfaces yet where the verb order differences between helix/vim come into play.
Then the question is readline. Not saying that vi or emacs deserve to win readline, but it’s up to you to describe how Helix mode would differ from vi mode.
Helix is wonderful. I did make one keybinding change: Switching in and out of Insert mode is CTRL-i so I don't have to wander off to the escape key so often.
>If only there were a more gentle path to editor modes. Maybe some simple graphical representation of the modes and commands that could be down in the corner?
I taught myself vim by setting the vim cheat sheet to be my terminal background, though I greyed it out slightly so it didn't obscure what I was typing. Once you have that there are only a few phrases you really need to know: "+p, "+y, "+d to access the system clipboard, :split and :vsplit, C-w w to switch windows, and g C-g for word/char count.
If I had a lot more free time, I've been noodling with some designs for a text editor with modes. It would start up by default in the equivalent of vim's insert mode where all the normal CUA keybindings (e.g. ctrl-c for copy) worked. But with many more. Then, if you go to the equivalent of normal mode you can just press X to do the same thing ctrl-X does in insert mode.
Seriously, why don't they just dedicate a group to creating the best pytorch backend possible? Proving it there will gain researcher traction and prove that their hardware is worth porting the other stuff over to.
reply