Difference between revisions of "User Agent"
From MgmtWiki
(→References) |
(→Solutions) |
||
| (14 intermediate revisions by the same user not shown) | |||
| Line 4: | Line 4: | ||
==Context== | ==Context== | ||
| − | *The | + | * The most useful [[User Agent]] on an internet connected device is a [[User]] [[Trusted Browser]] to work only in the user's best interests. The common browsers from Apple and Chrome try to pretend that's what they are, but the typically are designed to advance the interest in the manufacturer of monetizing their investment in the browser code. |
| − | *When discussing the use of the internet by a user, what is really meant is the presence of the user's agent on an internet connection. | + | * When discussing the use of the internet by a user, what is really meant is the presence of the user's agent on an internet connection which can process the content and [[JavaScript]] provided by the web site. |
| − | *Typically the [[Native App]] will run on a [[User Device]] in the user's physical possession, but it is possible that the agent is running on a service in [[The Cloud]]. Usually this would not be detected by a [[Web Site]]. | + | * Typically, the [[Native App]] will run on a [[User Device]] in the user's physical possession, but it is possible that the agent is running on a service in [[The Cloud]]. Usually this would not be detected by a [[Web Site]]. |
| + | * Most people would appreciate a [[User Agent]] that would track their activity and their preferences, but generally recoil at the idea that such personal information can be purchased on the open web. But they have been groomed to expect this to come at no cost to themselves. | ||
| + | * The most intriguing [[Native App]] is the [[Wallet]] which offers the promise of protecting user secrets and User Private Information and only releasing them as the user desires. | ||
| + | * For details on creation of a [[Native User Agent]] that really can be trusted by the user see the wiki page [[Native User Agent]]. | ||
==Problems== | ==Problems== | ||
| Line 18: | Line 21: | ||
*The most common way for [[User]]s to access a [[Web Site]] is with a web browser from some well-know browser provider either within a device operated by the user, or on a cloud computer under user control. | *The most common way for [[User]]s to access a [[Web Site]] is with a web browser from some well-know browser provider either within a device operated by the user, or on a cloud computer under user control. | ||
*While current web browsers do allow [[Web Site]]s to include programs as JavaScript to run within the browser, they do provide a very restricted [[Sandbox]] where the script must run for security reasons. | *While current web browsers do allow [[Web Site]]s to include programs as JavaScript to run within the browser, they do provide a very restricted [[Sandbox]] where the script must run for security reasons. | ||
| + | ===Digital Twins=== | ||
| + | Digital Twins<ref>Marco Furini +6, ''Digital Twins and Artificial Intelligence'' '''CACM 95''' no 4. (2202-04) p 98 ff</ref> can be used in learning environment to keep students engaged and on track. Of course, they can be incredible nags as well. | ||
| + | ===AI Assistants=== | ||
| + | In the short term, starting in 2024, the goal could be an AI Assistant that would have learned something from a LLM, but was augmented by what the user has actually done online. | ||
| + | * [https://www.wired.com/story/humans-forget-ai-assistants-remember-everything/ Humans Forget. AI Assistants Will Remember Everything] Digital assistants with access to everything you do on your phone can use that knowledge to remember the life details you’d otherwise forget. But with that great power comes great responsibility. | ||
==References== | ==References== | ||
| Line 24: | Line 32: | ||
[[Category:Glossary]] | [[Category:Glossary]] | ||
[[Category:Agent]] | [[Category:Agent]] | ||
| + | [[Category: User Agent]] | ||
| + | [[Category: Artificial Intelligence]] | ||
Latest revision as of 20:53, 12 April 2024
Contents
Full Title or Meme
A digital Entity that is operating at the consent and authority of the user.
Context
- The most useful User Agent on an internet connected device is a User Trusted Browser to work only in the user's best interests. The common browsers from Apple and Chrome try to pretend that's what they are, but the typically are designed to advance the interest in the manufacturer of monetizing their investment in the browser code.
- When discussing the use of the internet by a user, what is really meant is the presence of the user's agent on an internet connection which can process the content and JavaScript provided by the web site.
- Typically, the Native App will run on a User Device in the user's physical possession, but it is possible that the agent is running on a service in The Cloud. Usually this would not be detected by a Web Site.
- Most people would appreciate a User Agent that would track their activity and their preferences, but generally recoil at the idea that such personal information can be purchased on the open web. But they have been groomed to expect this to come at no cost to themselves.
- The most intriguing Native App is the Wallet which offers the promise of protecting user secrets and User Private Information and only releasing them as the user desires.
- For details on creation of a Native User Agent that really can be trusted by the user see the wiki page Native User Agent.
Problems
- Nearly any application running on a user's device is allowed to access the internet and claim that it represents the user. There is no built-in mechanism to test this assertion by an internet connected service. The internet was designed to connect computer systems, and that is all it can be relied upon to do.
- Any Web Site that wishes to create a Persistent Identifier for a User will need to take responsibility for any necessary Assurance that the program running on the user's device really does reflect the will of the user.
- Most of the larger enterprises operating on The Web prefer to supply a Native App to the users device to improve the User Experience for that site.
- To be sure that the User Agent really is operating on the user's behalf, the Web Site needs to know the level of Assurance that can be assigned to the user's device as well as the User Agent.
- Nearly every browser shipped lies their User Agent String to get the most web sites to accept them. The problem is that when they have different characteristics it is hard for the Web Site to determine which characteristics to use.
Solutions
- The most common way for Users to access a Web Site is with a web browser from some well-know browser provider either within a device operated by the user, or on a cloud computer under user control.
- While current web browsers do allow Web Sites to include programs as JavaScript to run within the browser, they do provide a very restricted Sandbox where the script must run for security reasons.
Digital Twins
Digital Twins[1] can be used in learning environment to keep students engaged and on track. Of course, they can be incredible nags as well.
AI Assistants
In the short term, starting in 2024, the goal could be an AI Assistant that would have learned something from a LLM, but was augmented by what the user has actually done online.
- Humans Forget. AI Assistants Will Remember Everything Digital assistants with access to everything you do on your phone can use that knowledge to remember the life details you’d otherwise forget. But with that great power comes great responsibility.
References
- ↑ Marco Furini +6, Digital Twins and Artificial Intelligence CACM 95 no 4. (2202-04) p 98 ff
