flash 8 pro is awesome... the last version by macromedia before adobe bought them.... still have my physical copy with serial ... it's a great little tool
no... things like inkscape/scribus take the place of the vector editing... and action script is just a limited version of JavaScript. It was excellent tool for building stuff back in the day, but now you can do everything with html/css/svg/js/es/ffmpeg/blender. I suppose if one wanted to prototype something it could still be useful... and I imagine it could still be used for it's creators original design, even today... frame by frame timeline vector animations. Flash wasn't originally designed to be a web technology.... but it did end up shaping it by letting ppl build whatever web pages their imagination could think of before doing so was possible via standard web browser technologies. So in a way it provided the canvas to envision the technology needed to build web applications as we know them today, but I don't actually use it to build websites anymore.
I remember flash animations and how easy to use and how many artists were using it around the time (2002?). I just asked because a lot of people don't want to learn all that stuff but everyone's scared of the flash is insecure warning (including me)
ah yes I remember now, how flash got killed off by the big browsers because of security flaws. I remember something about how Adobe wouldn't fix them at the time for whatever reason. But yes around 2002 it was popular.... it's downfall was a bit later however. You could build stuff that looked very cutting edge and didn't take much skill.... and stand alone browsers couldn't create that type of content yet so it was very popular. I wonder if anyone still creates web content with it today. It's been so long since I've seen it online that I forgot all about that insecure warning that appeared whenever a site embedded a swf file.
It was apple actually. Iphone wouldn't support flash. He hated it. The browsers went along with it. It was a battery killer so I didn't mind but I didn't know about it from the creator side. https://youtube.com/watch?v=7r7B_OqDIlM
ohh thats right! it was definetly apple that killed flash 100%.. forgot that detail. Thats not the only skelton in apples closet. Capacitive touch screens changed everything. RIP Palm Pilot and every other potential tech that could have become the norm but didn't. Ultimately, the world was given what the winners in big tech dished out.. which at their core are data harvesting nodes. I guess it was worth it though, the world has AI now, which is a pipe dream, and a century in development. Question is, what will the world look like in 20 or 30 years from now, knowing that many ideas never see the light of day. And what makes sense isn't what makes dollars.
I'd call what we have interesting pattern recognition that's not focused or certain enough. We have advanced markov chains for text, interesting text to image generation but I'm sure specialized computations like for Nvidia chip making is probably at least somewhat legit. Maybe Intel or AMD may be on top. I wonder what we'll see from the past in 20-30 years!
Their models might be good but their developer experience sucks ass compared to OpenAI and Anthropic. There's two very different systems for integrating with them (Vertex and the Gemini API) which also have different ways of doing things. The docs often don't make it clear sometimes what's being talked about until you are someway in. It's an utter mess. We have thousands of dollars of Gemini credit to spend and we've really struggled to get team members to use it.
That's before we even start talking about the utter clusterfuck that is paying as a _consumer_. It's like they looked at Microsoft's approach to bundling and thought it was a great idea, not realising you do that when you have a monopoly - not when you are trying to compete with a fast moving incumbent.
[You can use Gemini with the OpenAI library](https://ai.google.dev/gemini-api/docs/openai)...In that sense, I don't see how it could be worse than the other two.
You've fallen in to a trap and this is the kind of thing I'm talking about. That integration only works if you are able to use the Gemini API - which is only available for some people. For people with organisational accounts that don't have Gemini enabled (so most places) you have to use Vertex instead, which doesn't have an OpenAI-compatible endpoint.
I guess i'm a bit confused. I've had a GCP account for a few years now. It's just a free personal account that anyone can create. When I wanted to try out the Gemini API, all I had to do was generate an API key through Google AI Studio using the "Create API key" button. It seems like API keys can also be created within GCP if I didn't want to interact with Google AI Studio for whatever reason.
I guess the part that is confusing me is why Vertex needs to be involved instead of just enabling the Gemini API. I'm not too familiar with Vertex tbh. I also wonder if having prior experience with GCP and already having an account made the process simpler for me. I've heard other people complain about GCP, so my guess is you aren't alone. Personally, I use AWS at work now, and I would much prefer to work in GCP.
Anyway, that first comment was just to give you a heads up in case you didn't know about Gemini being compatible with the OpenAI library. I thought it was pretty convenient when I first discovered it.
> Anyway, that first comment was just to give you a heads up in case you didn't know about Gemini being compatible with the OpenAI library.
Thanks
> When I wanted to try out the Gemini API, all I had to do was generate an API key through Google AI Studio using the "Create API key" button. It seems like API keys can also be created within GCP if I didn't want to interact with Google AI Studio for whatever reason.
This only works if your Google account has AI Studio enabled, which I think is opt-in for organisational accounts (and requires an organisation admin to do so). In our organisation for example we have access to Vertex (via GCP) but not AI Studio.
The integration is already moving along nicely and provides a good experience even if the models aren’t as advanced. Drive + notebooklm + context window makes for a good experience. Not to mention all the free stuff
i’ve been using 2.0 flash in goog ai studio for awhile, it’s really good, and seems slop-free. pairing it with maps would have interesting logistics use cases
I'm going to test it out right now to help uncover vampire nests. Early Bing, back when it was Bing, was uncanny in its ability to suggest vampire hiding spots.
I’ve been using it for coding via api in open router and it keeps switching to speaking in Chinese, Hebrew, or Arabic. Even when it does work though, Claude 3.5 works much better.
Gemini just needs direct speech understanding so it can work better. Currently it works through a speech-to-text interpreter and then works on the text given, but it's not able to understand when words sound similar and get mis-transcribed, or if non-English words are used in an English sentence.
37 Comments
createthiscom@reddit
RouteGuru@reddit
InsideYork@reddit
RouteGuru@reddit
InsideYork@reddit
RouteGuru@reddit
InsideYork@reddit
RouteGuru@reddit
InsideYork@reddit
latestagecapitalist@reddit
Saffron4609@reddit
MaxDPS@reddit
Saffron4609@reddit
MaxDPS@reddit
Saffron4609@reddit
Pro-editor-1105@reddit
kayk1@reddit
West_League1850@reddit
BurritoOverflow@reddit
AnomalyNexus@reddit
epSos-DE@reddit
epSos-DE@reddit
MigorRortis96@reddit
nullmove@reddit
mrbluesneeze@reddit
GradatimRecovery@reddit
alcalde@reddit
Jumper775-2@reddit
GradatimRecovery@reddit
FerLuisxd@reddit
shouryannikam@reddit
chronocapybara@reddit
ZeeRa2007@reddit
Ulterior-Motive_@reddit
MixtureOfAmateurs@reddit (OP)
Beneficial-Good660@reddit
MixtureOfAmateurs@reddit (OP)