#statisticalengines — Public Fediverse posts

Read an interesting development of #AI use in drone #warfare operations and a confirmation from the #Ukrainian mililtary that fully autonomous #AI-Driven #drones is not the goal at all in the changing environment of the battlefield.

“ Much of the conversation surrounding artificial intelligence assumes the future belongs to fully autonomous drones. Battlefield experience suggests something more nuanced: soldiers prefer to be in the loop, because fully AI-enabled systems can actually make friendly forces less safe. In actual combat, over sustained periods, fully autonomous systems have trouble distinguishing friendly drones from foe. That confusion can cost battlefield advantage — and lives.

The objective therefore is not to remove humans from the fight; it is to remove unnecessary workload while preserving human judgment where it matters most. Recent Ukrainian interceptor systems automate much of the interception process, yet the operator still selects the target and retains the ability to abort the engagement. Even Ukraine’s most advanced autonomy initiatives preserve meaningful human control over the mission.

Which points to the more interesting question. It is not “how quickly can we take humans out of the loop?” It is “how do we build systems that military professionals actually trust when the shooting starts?” Trust — not autonomy — is likely to be the decisive factor in whether autonomous systems get adopted at all.” (Source: Breaking Defence)

I note that the #AI which is discussed here has little to do with the LLM-based GendAI slop machines peddled by #TechBros nor the massive #DataCentres these #StatisticalEngines require to produce #simulacra of text for human consumption.

#SlavaUkraini #DroneWarfare

Read an interesting development of #AI use in drone #warfare operations and a confirmation from the #Ukrainian mililtary that fully autonomous #AI-Driven #drones is not the goal at all in the changing environment of the battlefield.

“ Much of the conversation surrounding artificial intelligence assumes the future belongs to fully autonomous drones. Battlefield experience suggests something more nuanced: soldiers prefer to be in the loop, because fully AI-enabled systems can actually make friendly forces less safe. In actual combat, over sustained periods, fully autonomous systems have trouble distinguishing friendly drones from foe. That confusion can cost battlefield advantage — and lives.

The objective therefore is not to remove humans from the fight; it is to remove unnecessary workload while preserving human judgment where it matters most. Recent Ukrainian interceptor systems automate much of the interception process, yet the operator still selects the target and retains the ability to abort the engagement. Even Ukraine’s most advanced autonomy initiatives preserve meaningful human control over the mission.

Which points to the more interesting question. It is not “how quickly can we take humans out of the loop?” It is “how do we build systems that military professionals actually trust when the shooting starts?” Trust — not autonomy — is likely to be the decisive factor in whether autonomous systems get adopted at all.” (Source: Breaking Defence)

I note that the #AI which is discussed here has little to do with the LLM-based GendAI slop machines peddled by #TechBros nor the massive #DataCentres these #StatisticalEngines require to produce #simulacra of text for human consumption.

#SlavaUkraini #DroneWarfare

Read an interesting development of #AI use in drone #warfare operations and a confirmation from the #Ukrainian mililtary that fully autonomous #AI-Driven #drones is not the goal at all in the changing environment of the battlefield.

“ Much of the conversation surrounding artificial intelligence assumes the future belongs to fully autonomous drones. Battlefield experience suggests something more nuanced: soldiers prefer to be in the loop, because fully AI-enabled systems can actually make friendly forces less safe. In actual combat, over sustained periods, fully autonomous systems have trouble distinguishing friendly drones from foe. That confusion can cost battlefield advantage — and lives.

The objective therefore is not to remove humans from the fight; it is to remove unnecessary workload while preserving human judgment where it matters most. Recent Ukrainian interceptor systems automate much of the interception process, yet the operator still selects the target and retains the ability to abort the engagement. Even Ukraine’s most advanced autonomy initiatives preserve meaningful human control over the mission.

Which points to the more interesting question. It is not “how quickly can we take humans out of the loop?” It is “how do we build systems that military professionals actually trust when the shooting starts?” Trust — not autonomy — is likely to be the decisive factor in whether autonomous systems get adopted at all.” (Source: Breaking Defence)

I note that the #AI which is discussed here has little to do with the LLM-based GendAI slop machines peddled by #TechBros nor the massive #DataCentres these #StatisticalEngines require to produce #simulacra of text for human consumption.

#SlavaUkraini #DroneWarfare

Read an interesting development of #AI use in drone #warfare operations and a confirmation from the #Ukrainian mililtary that fully autonomous #AI-Driven #drones is not the goal at all in the changing environment of the battlefield.

“ Much of the conversation surrounding artificial intelligence assumes the future belongs to fully autonomous drones. Battlefield experience suggests something more nuanced: soldiers prefer to be in the loop, because fully AI-enabled systems can actually make friendly forces less safe. In actual combat, over sustained periods, fully autonomous systems have trouble distinguishing friendly drones from foe. That confusion can cost battlefield advantage — and lives.

The objective therefore is not to remove humans from the fight; it is to remove unnecessary workload while preserving human judgment where it matters most. Recent Ukrainian interceptor systems automate much of the interception process, yet the operator still selects the target and retains the ability to abort the engagement. Even Ukraine’s most advanced autonomy initiatives preserve meaningful human control over the mission.

Which points to the more interesting question. It is not “how quickly can we take humans out of the loop?” It is “how do we build systems that military professionals actually trust when the shooting starts?” Trust — not autonomy — is likely to be the decisive factor in whether autonomous systems get adopted at all.” (Source: Breaking Defence)

I note that the #AI which is discussed here has little to do with the LLM-based GendAI slop machines peddled by #TechBros nor the massive #DataCentres these #StatisticalEngines require to produce #simulacra of text for human consumption.

#SlavaUkraini #DroneWarfare

Read an interesting development of #AI use in drone #warfare operations and a confirmation from the #Ukrainian mililtary that fully autonomous #AI-Driven #drones is not the goal at all in the changing environment of the battlefield.

“ Much of the conversation surrounding artificial intelligence assumes the future belongs to fully autonomous drones. Battlefield experience suggests something more nuanced: soldiers prefer to be in the loop, because fully AI-enabled systems can actually make friendly forces less safe. In actual combat, over sustained periods, fully autonomous systems have trouble distinguishing friendly drones from foe. That confusion can cost battlefield advantage — and lives.

The objective therefore is not to remove humans from the fight; it is to remove unnecessary workload while preserving human judgment where it matters most. Recent Ukrainian interceptor systems automate much of the interception process, yet the operator still selects the target and retains the ability to abort the engagement. Even Ukraine’s most advanced autonomy initiatives preserve meaningful human control over the mission.

Which points to the more interesting question. It is not “how quickly can we take humans out of the loop?” It is “how do we build systems that military professionals actually trust when the shooting starts?” Trust — not autonomy — is likely to be the decisive factor in whether autonomous systems get adopted at all.” (Source: Breaking Defence)

I note that the #AI which is discussed here has little to do with the LLM-based GendAI slop machines peddled by #TechBros nor the massive #DataCentres these #StatisticalEngines require to produce #simulacra of text for human consumption.

#SlavaUkraini #DroneWarfare