71

Haptek / Re: Haptek rises from the ashes

« Last post by squarebear on May 17, 2024, 11:34:19 am »

Hi 8pla. Sorry, I meant that the bot itself doesn't respond to any input. Those messages are from the website rather than the bot. As the company no longer seems to exist, I doubt that it will be fixed, but I admire your optimisim 

72

Haptek / Re: Haptek rises from the ashes

« Last post by 8pla.net on May 17, 2024, 08:58:45 am »

Hello there Squarebear!

Great topic.... Let's chat more about it!

The demo is still online (Chrome only), but the bot no longer responds to input.
https://virtyou.com/mitsy/m.html

Mitsy in Firefox responds to the Start button input with, "sorry. i can't hear you unless we are in a chrome browser" and then, "ok, enough is enough"

Hi Freddy!

I also discovered a few more ashes in the Web Developer Tools menu: 

Mitsy also says, "don't worry. will fix it in the next version. but for now, just use chrome. unless you wanna just do the eye contact thing"  and "come on. switch to chrome so we can blab. ok?"


BREAKING NEWS!
They will fix Mitsuku in the next version

Citation:
Virtyou. (n.d.). Mitsy [Webpage]. Retrieved from https://virtyou.com/mitsy/m.html

73

General Project Discussion / Re: Reasoner.js: a framework for generalized theory synthesis

« Last post by ivan.moony on May 15, 2024, 12:04:13 am »

I had some experiments lately with symbolic algorithm synthesis. I had an input, and I had an output, and I had to clue up how can I brute force combine given rules to transform the input to the output. It works well with a small amount of steps, but eventually combinatorial explosion happens where I get stuck in almost endless loop that may take centuries to finally get an answer.

And that is where neural networks seem to do very good job: to quickly learn how to connect inputs to outputs, memorizing weights for future construction of outputs. Of course, there is an approximation error with neural networks, but given enough learning examples, it may work like a charm.

Basically, the two technologies should do the same thing, but with different advantages and drawbacks. The symbolic one brings certainty but it is slow while the neural one is fast but it brings randomness. Is combining them really the next big thing the world is waiting for?

74

AI News / Re: Open AI GPT-4o - audio, vision, text combined reasoning

« Last post by MikeB on May 14, 2024, 05:46:48 am »

They are actually claiming reduced tokens in language. "English 1.1x fewer tokens (from 27 to 24)", with some languages compressing more.

Although there's little to say how it works, as Noam Chomsky points out, nothing is learned from language if a computer arbitrarily uses language and maths to predict with, including arbitrary pre-compression...

Response times are good however.

ChatGPT 3.5 is fast as well. https://chatgpt.com/

75

AI News / Open AI GPT-4o - audio, vision, text combined reasoning

« Last post by MikeB on May 14, 2024, 05:37:05 am »

GPT-4o (“o” for “omni”) is a step towards much more natural human-computer interaction—it accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. It can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds, which is similar to human response time(opens in a new window) in a conversation. It matches GPT-4 Turbo performance on text in English and code, with significant improvement on text in non-English languages, while also being much faster and 50% cheaper in the API. GPT-4o is especially better at vision and audio understanding compared to existing models.

https://openai.com/index/hello-gpt-4o/

76

General Hardware Talk / Butler In A Box (1983)

« Last post by MikeB on May 09, 2024, 09:39:06 am »

This is a fun video about Butler In A Box, sold from 1983-2000.

It was sold as voice recognition hardware (for up to 4 different users) that could respond as a bulter and turn on different electric items in the house. Lamp, TV, etc.

It had a huge price tag and notoriety for two decades in futuristic home wares, but does not have one mention in wikipedia as a standout in AI/voice recognition.

77

AI News / AI controlled F-16, for real!

« Last post by frankinstien on May 04, 2024, 01:04:11 am »

Interesting article and the implementation of 1,000 aircraft controlled by AI is the tipping scale. Those nations that can harness AI don't need to sacrifice lives to win wars. Welcome to the game of World Domination, or more like the 1983 movie WarGames: "Shall We Play a Game?"   

78

Haptek / Re: Haptek rises from the ashes

« Last post by Freddy on May 03, 2024, 01:28:56 pm »

Ahh I see, thanks for the update 

79

Haptek / Re: Haptek rises from the ashes

« Last post by squarebear on May 03, 2024, 07:53:52 am »

I wouldn't get too excited about this. Virtyou hasn't been active for many years. We partnered with them to make an avatar for Mitsuku about 7 or 8 years ago. The demo is still online (Chrome only), but the bot no longer responds to input. Shame really, as it was really cool when it was working and great to demo on a phone.
https://virtyou.com/mitsy/m.html

Virtyou.ai later became expressive.ai but as their website is no longer online, I guess they don't exist any more. However, there is an old Youtube video from Mobilebeat 2017, which shows Angela Ford (the CTO of Expressive.ai) presenting a demo of Mitsuku and the avatar: https://youtu.be/UljiivP9aVo?si=4dky_-_jtevrjJeF&t=166

80

Haptek / Haptek rises from the ashes

« Last post by Freddy on May 03, 2024, 12:27:21 am »

There is something interesting at a redirect from the old Haptek domain!

https://virtyou.com