We all have some confirmation bias, when that bias passes a certain level it becomes wishful thinking also known as religion.
One day the Christian God comes back to earth. Many question his legitimacy, and he gives enough proof that he is the one he claims to be. At that point, there is no doubt in the public mind that he is God. The questions pile up and apparently, there is no end to this. Meanwhile, the people begin to form three distinctive groups:
1. Most of the people are in denial. They accept the fact but do not agree with it. For them, religion is not only about God but identity, values, tradition, culture, and overall a big enough part of their lives to be changed by whoever (even by Him). For them, the reality always has been a matter of personal choice. They choose what makes them happy and what could be more important to a person.
2. The second group we may call opportunists. They accept the fact, leaving their feelings for later. They try to position themselves in the context of new reality so they can benefit somehow. I’m not talking only about material gain, there are plenty of things in their wishlists – having power, being famous, having more time, etc. New business books are written, like “How to make God work for you?”. Televangelists (tele-evangelical pastors) are first in line.
3. The third group is more decisive (and extreme) than the others. Once they are certain to deal with the real God, the idea that God has no place on earth. God is part of an enormous socio-phycological environment that functions on the premise God is mysterious and mostly inaccessible (not here). As a result, the only logical conclusion is to send God back to wherever He came from. Not necessarily by crucifying him one more time but with the same outcome.
“Credo quia absurdum” or “I believe because it is absurd” is one of the pillars of the Christian faith.
Consciousness is a millennia-long debate and probably will go on till the end of our species. We disagree on almost everything about consciousness and professionals dealing with consciousness avoid the term or use it in a very narrow sense (as the sensation of having experiences). We cannot provide a general definition of consciousness to the point that we don’t agree on what a definition of consciousness should look like.
However, there is one practical aspect we face about consciousness: conscious machines and it is mostly answering two questions:
- should we allow our computers to reach a level of complexity and autonomy that one way or another will lead to the emergence of something which is consciousness in humans?
- in the case of conscious AGI, how are we supposed to treat it? It is not a moral question (at least at first) but a practical one. Conscious AGI will have an initial set of values and motivations but inevitably it will evolve by modifying its code. The only way to coexist with conscious AGI is to recognize some common rights analog to human rights.
Why do LLMs however useful they are in many domains, are considered a cul-de-sac of AI development?
Let me try to simplify it for you. Imagine you have a sine function represented by a number of points and you attempt to model that function with polynomes. You can increase the number of points (more data) and you may increase the polynomial degree (better models) but you stay within the polynomial model.
What will happen: your model will work better and better for interpolation and short-range extrapolation but you will never get to the point that you can predict behavior far beyond the range of the training set. That is LLMs simplified.
The solution is to make AI use more than a polynomial model, which is the direction AI is heading right now. The ability of AGI to create models unknown to the AGI from history. Depending on your school of thought, you may hear terms like reasoning, hierarchical planning, or meta-learning, not the same thing but with the same aim.
Exciting times!
Imagine you’re a parent of a superhero. Your child will be able to save or destroy the planet when (s)he comes of age. Can you guarantee that you will be able to raise a good person or at least not an evil one? After you educate yourself and put in all your efforts you can only hope…
Now imagine that there are many like you and your child. Some of them share your humanistic values but others think that the glory of god or some ideology are above human life. What are we going to do about them, ship them to Mars?
All my life I have been thinking that in case someone offered me time travel I would go to the future. Not too far, say 50 to 100 years, so I can recognize the surroundings. Now I’m thinking 5 to 20 years would be enough. In 5 years I may wish to go back…
Scripthea is a prompt engineering tool that offers a systematic approach to composing prompts for text-to-image AI generators. The intended usages vary from familiarizing yourself with the basics of assembling a prompt via creating and stabilizing your own style to routine image creation for highly trained individuals.
Prompts in Scripthea are made up of cues and modifiers. A cue is a short descriptive text of the image, while modifiers can be attributes like a painter, an art style, or a time period. After composing a prompt, users can either copy and paste it into their preferred text-to-image generator or use API access to a local installation of Stable Diffusion, Automatic1111, or ComfyUI distributions. In the former case, all the options of the Stable Difussion web interface are available for adjustment.
The software provides two modes for prompt composition: Single and Scan. In Single mode, a user can use one cue and one or more modifiers. In Scan mode, any number of cues and any number of modifiers can be selected, but each prompt will combine only one cue with a number of modifiers following some rules. Modifiers are organized into checkable categories. The user should specify the image depot folder where the generated images will go.
Scripthea also includes an image depot viewer that shows a Scripthea image depot. An image depot is a folder with a bunch of images and a “description.idf” file for the prompts and metadata. The viewer has options for table view and thumbnail view, and in the grid view, the thumbnails can be adjusted from a menu. It also has a find panel that can search for words in the prompts of the active image depot.
The third tab is image depot master which is a manager for image depots. It implements operations like a copy, move, or delete of selected subsets of images in a familiar table or grid view.
The fourth tab of Scripthea contains a utility for image files from external generators (like craiyon.com). The Scripthea utility will import these images into an image depot. An export utility within the same tab will export a selection from an image depot to a new folder with an optional web page for easy visualization outside Scripthea.
A recent addition is an integrated Python scripter for an automation of routine tasks and/or open Scrtipthea to external Python software.
For any assistance, pressing F1 will lead you to online help.
Scripthea software was created by Teodor Krastev in C# with WPF and is distributed under an MIT open-source license. Users are encouraged to report any bugs or provide suggestions for improvements using the contact link at Scripthea.com.