Solextre

Solextre

@solextre.bsky.social

Retired teacher (systems theory, communication) maker of abstract music videos (see solextre YouTube) no political party ... to quote Kurt Vonnegut: "To be is to do - Socrates To do is to be - Sartre Do Be Do Be Do - Sinatra" :-)

499 followers50 following3.9K posts

#CHRNCL "Once we have surrendered our senses and nervous systems to the private manipulation of those who would try to benefit from taking a lease on our eyes and ears and nerves, we don’t really have any rights left." [McLuhan]

#CHRNCL It is hard to grasp the power of this. For example, a political issue may be discussed by thousands of groups which then reform to discuss the same issue from a different point of view. This comes close to being true democracy, something our current representative system cannot reproduce.

#CHRNCL These collections of nodes share information. Thus, more than at any previous time in human history, each of us knows what everybody else is thinking. This is not a one-to-one understanding of course, that would be impossible – rather it exists as an global impression.

#CHRNCL In this way, our phones become links in a chain of experiences that may possibly be giving rise to a global culture. It is worth noting that Chinese software – TikTok for example – tends to use a similar organization...most likely because Chinese software was developed from American models.

#CHRNCL Regardless of the human languages involved, these media generate cultural change, and the organization of the software, and the tasks that it can automatically perform – such as texting, photography, music, and so on – reflect the culture of the software's creators – mostly the Americans.

#CHRNCL This software automatically translates itself into the language of the person using the phone – for example if you live in Germany and by a Samsung phone running the Android operating system, all of the text on phone will be in German, and the assistant will speak German.

#CHRNCL The new technologies also have cultural implications. There are approximately 9 billion smartphones in the world, most of them made by Samsung (South Korea) and Apple (United States). Samsung phones use software developed by Google, and Apple uses its own software - both American companies.

#CHRNCL The power of this feature has been already shown itself. For example, when ICE intruded into Los Angeles and Minneapolis, protesters use their phones to organize demonstrations. These demonstrating groups had never existed before, and after ICE left town, the groups still existed.

CHRNCL For "family", you can substitute any cohesive group of people. The side effect of the network is the sudden rise in the kinds and sizes of groups. In the Internet age an individual might belong to 20 or 30 different groups.

#CHRNCL For example, a family reunion involving 30 or 40 people might be planned by a small group using a social media app to set up a calendar, and list of who will bring what to the party, while allowing all members to review what is going on and make comments.

#CHRNCL Similar subsets exist in many families. Although parents, children, and grandchildren may be spread out across the country, they remain connected via their phones. Furthermore this is a very rich connection, supporting videos, photos and calendars among others.

#CHRNCL: Various subsets of the global communications network exist permanently. For example, Bluesky is a network of phone nodes. We call it "social media", but that is just a convenient verbal tag. Bluesky exists as a collection of individuals who are connected to each other via the Internet.

#CHRNCL: The Global Village "As electrically contracted, the globe is no more than a village. Electric speed in bringing all social and political functions together in a sudden implosion has heightened human awareness of responsibility to an intense degree." [McLuhan]

#CHRNCL To gamble is to make a guess. Prediction markets are used by people who have access to the best data and the most effective information tools. Anyone who tries to guess in a prediction market will be wrong most of the time. Conclusion: prediction markets are not gambling.

#CHRNCL It means, for example, that a group prediction that a forest fire will burn for two more weeks suggests that the information on hand supports the prediction. Based on this the fire companies can confidently plan to be in the field for that period of time.

#CHRNCL Prediction markets do have one very interesting positive benefit - they measure belief. If if you are willing to back your assessment with money, then you believe that you are correct. This is a more accurate way to measure what people actually think than are opinion polls.

#CHRNCL It is not clear that the problem has a solution. In fact, as more and more of our economic and cultural activities depends on the sharing of data, it is likely that the use of information for personal advantage will increase. Laws to prohibit this have never been very effective.

#SCIX When you go to a McDonald's now and order at the drive-through, you are talking to their new AI ordering system. According to McDonald's, the AI understands a higher percentage of orders than the human employees that it replaced. (They have been assigned to other jobs inside the restaurant.)

#CHRNCL In poker, "insider information" would be knowledge of the other players' cards and would be cheating. But in a prediction market, "insider information" is simply knowledge of the world. A society may decide to keep some information secret, but that is a different sphere of influence.

#CHRNCL There is much more to be said about AI, but this is not the forum for that discussion. If you are interested in going deeper, I would suggest searching "AI" at YouTube. Next: a return to "prediction markets".

#SCIX Human language is highly metaphorical. When Model-T Fords began appearing on the roads, a word was needed to describe their power. People chose "horsepower" – because horses were how they got around. But that doesn't mean that a 300 HP engine is the same as a wagon pulled by 300 horses.

View full thread

So it's fair to say that an AI executes a program, stores the results of that program into its database, and uses that data to re-execute the program in a new way. But this is neither "learning", nor "planning", nor "intelligence". Someday those might develop, but they are not here yet.

But AI IS something UTTERLY unique in human experience – computer software that can perform complex tasks without supervision. We should not attach human terminology to this, but we should recognize that we have invented an extremely powerful new tool, and that we need to learn how to use it.

#SCIX Human language is highly metaphorical. When Model-T Fords began appearing on the roads, a word was needed to describe their power. People chose "horsepower" – because horses were how they got around. But that doesn't mean that a 300 HP engine is the same as a wagon pulled by 300 horses.

We are doing the same thing with AI. We say that AI "learns". But AI is not alive, it has no experience, and it has no way of "remembering" its past. What it has is a storage area into which bits of data can be placed. This gives it the ability to reuse data, but it does not "learn".

Similarly, people often talk about AI activities as if the AI were planning ahead and thinking through problems. But AI agents do not have thoughts. What they have is software routines that human programmers have designed to deal with certain situations. AI can act, but it doesn't think.

#SCIX Some examples. There already exist chemical and radiological treatments for cancer that would work extremely well if there was a way to deliver them to the tumors. Nanobots piloted by a cohort of AI agents could provide a straightforward way of delivering these medications.

View full thread

A city's traffic lights, in conjunction with cameras located at key intersections, could provide data to an AI cohort to coordinate traffic flow across an entire urban area. Current systems that attempt to do this are limited by computers that are unable to process the large amount of data required.

Almost any large scale logistic system could be coordinated on a global basis by means of an AI cohort associated with satellite data transmission from transportation vehicles. <<<And there are many, many more examples.>>>

#SCIX Some examples. There already exist chemical and radiological treatments for cancer that would work extremely well if there was a way to deliver them to the tumors. Nanobots piloted by a cohort of AI agents could provide a straightforward way of delivering these medications.

When drones are used as weapon systems, each drone must be piloted by a human. Assigning an AI cohort to pilot a group of drones could essentially automate the use of this weapon system. This idea applies to cruise missiles, artillery, and other kinds of projectile weapons.

A large AI cohort given access to the nation's air traffic control systems could coordinate air traffic across the entire country. Today this control is only established within a certain distance of each particular airport.

#SCIX The two most obvious problems mentioned earlier: accidental damage caused by AI agents - and intentional damage caused by people using AI agents. Until the OpenAI event, no one knew that agents could work together in large "swarms". This fact will now be investigated, extended, and put to use.

View full thread

Nations like Russia, Ukraine, US, China and others are already developing advanced technologies, such as drones, as weapons. There is almost no doubt that AI agent swarms will be developed similarly. Terrorist and criminal organizations will also find uses for the AI agent technologies.

In a nutshell, we have invented a technology that can be used as a weapon, and we need to develop defenses that will prevent such attacks. This needs to be done by everybody, right now, on a global scale. That is why so many people are talking about this event.

#SCIX The two most obvious problems mentioned earlier: accidental damage caused by AI agents - and intentional damage caused by people using AI agents. Until the OpenAI event, no one knew that agents could work together in large "swarms". This fact will now be investigated, extended, and put to use.

In the rush to develop new applications, prototype AI swarms may find their way into the global Internet. AI agents run on computers. In looking for computers to host their work, agents may accidentally cause crashes in systems anywhere in the world.

This may not seem all that serious, but such systems might include traffic control, hospitals, banks, phones and other parts of our electronic infrastructure that are essential to our daily lives. We can compensate for occasional breakdowns, but AI could make the situation chronic.