I think over the years local models have fairly consistently been ~7 months behind frontier performance. Local models are hugely important but I don’t see the calculus changing. I can imagine it’s certainly the case for many tasks that there will be diminishing gains for performance improvements or reliability pass some threshold, in which case you don’t need frontier performance and you can certainly use local models or at least cheaper tiers of proprietary models if local is too much of a hassle. Plus of course use cases where local is necessary or the pros of having local models or on device models outweighs that of frontier.
Maybe things will change though, I would assume through basically government subsidies from China etc, to undercut existing frontier labs, but you can always spend more (better data more compute etc) for better performance and that I can imagine will always have a selling point.
I read it that strong emotions have a signature that's processed differently in the context of data points vs the context of human interaction in meatspace.
It feels to me like an intro to a deeper insight that's missing the main body of the piece.
Yes, removing ambiguity can trim too much away. If we could describe what that is, it would not have been thrown away.
That is an interesting idea because we see this loss everywhere even in writing, photographs, etc. long before we ever considered "data points". It's such a profound loss that when an artist manages to capture even a tiny glimmer of emotional ambiguity they are hugely celebrated.
We are all starving for this while the world we build insists on throwing more and more of it away.
One of the things that I've noticed is there's an aversion to people noticing their emotions, companies analyzing that data, "men sharing their feelings" and the way that I've come to process these things is that at the end of the day they're all signals.
For example the gentlemen that we crossed. His rage, contempt, anger, disgust, whatever it may be, was a signal and the way that we interreacted had a profound impact on the other signals that came to surface.
I believe when we look at these signals from a functional perspective and truly consider the "subjective" nature of the signals, then it makes sense that they are some of the most important signals we have.
You put your hand in hot scolding water, most people immediately take note of the signal and act accordingly.
There's been a lot of devaluation on the actual supportive and objective nature of our own emotional and cognitive signals.
The mere idea that a corporation is tracking emotional data becomes absurd because at the end of the day, we're all kind of doing that ourselves. The tone of how we associate is a bit dysfunctional if you ask me.
I'm an adult with a fairly balanced view of AI and I find it difficult to learn coding without occasionally using AI to help me navigate to the most relevant bits of MDN or help me check if my thinking on an approach is correct (it's all entry level stuff so should be well represented in the training data).
I find it easy to to into a long chat with an LLM about some project I'd like to try and what's involved with it. I find it easy to get into a chat with an LLM about a lot of things as a kind of unproductive excursion that my brain tells me at the time is 'useful'. I'm of average will, so I dread to think how this will work out with children who get to 'partner up with AI to assist them' or whatever marketing speak is used to obfuscate their goals. Then combine that with social developmental issues or below average focus.
It's bleak because the more entangled they get with the system the more they'll seek to push back regulations.
My partner was working at an event and a co-worker had prepared a poster using AI - a teenage kid at the event pointed out how the poster "has AI smudges".
You know how your parents are weirdly shitty at recognizing obvious photoshops? Kids are constantly surprised that we adults can't recognize obvious AI images.
We use Teams at work and when you choose the icon it takes you to a screen that has a large icon of a door with a rope in front of it. From there you get to choose Teams on web or Teams the app. The point of the door is to tell you that Teams classic is no longer available, which is a huge part of the visual hierarchy. It's very strange - Teams classic was phased out long ago, but they still tell you this, and the negative connotation of a door with a rope in front of it resides in your mind as you move forward. This is one of the many operating quirks one sees from day to day.
I also took it that the benefit of being a good photographer and not using digital is that you take less photos and may end up more efficient as a result. I could be wrong but it's cumbersome when wedding photographers share up a 7GB folder of photos with previews that are 128 pix wide, within which there are runs of photos from the same angle in their dozens which require you to full screen load a 4000+ px wide image, to see which one has the best smile or whatever. It just feels like this is too much storage and admin to me.
I've often mused about how people get irritated by others being optimistic about change when the observers have tried change in the past and not been able to maintain it. I feel that the experience of that can lead to a position of cynicism that is defined by ones own limitations rather than the constraints of the system. They'll even suggest that people should be stronger in their resistance against the proven stickiness of platforms that use huge data to keep people in their ecosystems.
As someone who has eaten way too much sugary food I think my gut-brain coupling may have had enough of this. A few weeks ago I had a sugar binge one night and the cognitive effects were impossible to ignore the next day. Fortunately after 2-3 days I was back to normal but of my sample size of one, and in my condition (which is pre-diabetic) I observed a clear link.
It was a good experience as it's prompted me to get more serious about cutting back sugar, implemented as long term, achievable habit change.
It simply depends on what your needs are IMO - You can do great magazine design in Affinity, brochures, flyers, logos all that stuff. The only thing I'd miss in InDesign is image expand probably.
Also if you're making video games, and you don't need to export multi-res textures and work on the edge of file formats for advanced texturing etc, if your budget and needs are served by Affinity why spend on Adobe?
How far behind are models that can be run locally, and do you expect that this will be widespread?
reply