Can Artificial Intelligence Be Applied to Logistics?
Summary
TLDR在这段访谈中,美国军事大学在线的物流学教授奥利弗·海奇皮斯(Oliver Hedgepeth)讨论了人工智能(AI)在物流领域的应用。他提到自1985年以来,AI已经无处不在,从餐厅、家庭到军事和民用领域都有其身影。AI帮助士兵执行任务,机器人协助战斗,甚至在外科手术和餐饮业中发挥作用。海奇皮斯教授还提到了AI在物流中的具体应用,如亚马逊使用AI和机器人来优化订单处理和配送。尽管AI在物流中取得了巨大进步,但仍存在信任问题,人们对AI系统的偏见和可靠性持有疑问。他强调,尽管AI能够解决问题,但人类仍然需要对AI的决策进行判断和监督。
Takeaways
- 🧠 人工智能(AI)已经无处不在,从家庭到军事领域都有广泛应用。
- 🤖 机器人技术是AI的一个重要分支,军事和民用领域都有其身影。
- 🛒 亚马逊等公司在物流领域使用AI和机器人技术,提高了效率和准确性。
- 📈 AI在供应链管理中发挥作用,帮助追踪产品和解决延迟问题。
- 🔍 AI的定义随时间变化,现在有多种定义,取决于个人的观点和背景。
- 🛑 信任成为AI应用的一个新限制,人们对AI的信任程度不同。
- 🧐 人们对AI的偏见和背景会影响AI系统的构建和问题解决方式。
- 🎵 AI可以快速创造音乐效果,显示出在创造性任务中的潜力。
- 🤔 随着AI的发展,人们对于是否完全依赖AI的决策存在疑问。
- 🔑 人类判断在AI决策中仍然扮演重要角色,不能完全交给机器。
- 🚫 AI系统存在局限性,并非所有任务都能完美执行,如机器人引导逃生的失败案例。
Q & A
人工智能在物流领域的应用是如何开始的?
-人工智能在物流领域的应用始于1985年左右,从那时起就开始了对其应用的探索,至今已经广泛应用于各个领域。
奥利弗·海吉佩斯教授如何看待人工智能在物流中的作用?
-奥利弗·海吉佩斯教授认为人工智能在物流中的作用无处不在,它帮助士兵执行任务,也在民用领域,如餐厅、家庭和在线购物等方面发挥着重要作用。
人工智能的定义是如何随时间变化的?
-人工智能的定义随着时间的推移而发生了变化。最初,它旨在模仿人类大脑的思维方式,而现在它更多地被定义为日常使用的实用性技术。
机器人技术在人工智能中扮演什么角色?
-机器人技术是人工智能的一部分,它们在军事和民用领域都发挥着重要作用,如帮助战斗、侦察、携带物资等。
在物流领域,人工智能有哪些令人兴奋的应用?
-在物流领域,人工智能用于跟踪库存、优化供应链、自动化仓库管理、快速处理在线订单等,极大地提高了效率和准确性。
人工智能在物流应用中的限制是什么?
-人工智能在物流应用中的限制包括信任问题、偏见问题以及人类对完全自动化决策的接受程度。
为什么信任在人工智能应用中如此重要?
-信任对于人工智能至关重要,因为人们需要相信AI系统提供的解决方案是正确和可靠的,尤其是在关键决策和行动中。
人工智能系统是否存在偏见,如果是,它们是如何产生的?
-是的,人工智能系统可能存在偏见,这些偏见通常来源于创建系统的人的背景、经验和价值观,这可能导致AI系统在解决问题时产生不同的结果。
人类在人工智能决策中扮演什么角色?
-人类在人工智能决策中扮演着关键角色,他们需要对AI系统的输出进行判断和选择,以决定是否执行AI的建议。
人工智能在预测和规范性建议方面的未来发展方向是什么?
-人工智能正从描述性系统发展到预测性系统,最终可能发展为规范性系统,这意味着AI将能够提供具体的行动建议。
我们是否应该完全信任人工智能系统,并且不关心它是如何得出结论的?
-尽管人工智能系统在某些方面比人类更高效,但我们仍需要人类进行最终的判断和决策,以确保AI系统的建议是安全和合适的。
人工智能在紧急情况下的应用有哪些局限性和风险?
-在紧急情况下,如火灾逃生,人工智能可能无法完全可靠地指导人们安全撤离,这要求人们不能完全依赖AI,而需要保持警惕和自主判断。
Outlines
😀 人工智能在物流中的应用
本段落介绍了人工智能(AI)在物流领域中的应用。奥利弗·海吉佩斯教授,作为美国军事大学在线物流学教授及陆军人工智能物流中心的创始主任,分享了自1985年以来他在AI物流应用方面的经验。他指出AI无处不在,无论是在餐厅、家庭、军事还是民用领域都有广泛应用。AI帮助士兵执行任务,机器人和AI在医疗、餐饮等领域的应用也越来越广泛。此外,还讨论了AI的定义随时间的演变,以及人们对AI的接受程度和信任问题。
🤖 机器人技术与AI的结合
这一部分深入探讨了机器人技术在军事和民用物流领域的应用。军事领域中,机器人帮助作战,使用无人机侦察和打击敌人,以及在夜间帮助士兵避开地雷。民用领域中,AI和机器人在物流系统中扮演着重要角色,如亚马逊使用AI和机器人挑选和打包货物,以及在仓库管理中跟踪产品。AI还帮助处理由于疫情导致的供应链延迟问题。然而,AI在物流领域的应用也存在局限性,主要是人们对于AI和机器人的信任问题。
🔍 AI在物流中的局限性和人类的角色
在这一部分中,讨论了AI在物流中应用的当前局限性,特别是信任问题。美国国家标准技术研究所(NIST)的AI小组研究了人们对机器人和AI系统的信任,并探讨了AI的不同定义和类别。AI系统可能因为开发者的背景和偏见而不同,这影响了人们对AI的信任。尽管AI在解决问题方面取得了进展,但人类判断在执行AI系统的结论时仍然扮演着重要角色。此外,AI系统的决策过程是否透明,以及人们是否接受由“黑箱”直接输出的结论,也是值得关注的问题。
Mindmap
Keywords
💡人工智能
💡物流
💡军事应用
💡机器人
💡信任
💡偏见
💡供应链
💡预测系统
💡规范系统
💡黑箱
Highlights
人工智能(AI)在物流领域的应用非常广泛,包括军事和民用领域。
AI在日常生活中无处不在,如餐厅、家庭、军事和民用领域。
AI帮助士兵执行任务,如机器人和无人机的使用。
AI在民用领域也有广泛应用,如亚马逊物流系统。
AI在物流中用于跟踪产品,应对供应链延迟问题。
AI的定义随时间变化,不同人有不同的理解。
机器人技术在军事和民用领域同样无处不在。
AI在物流中用于解决复杂问题,如供应链跟踪。
AI的局限性之一是人们对其的信任问题。
美国国家标准技术研究院(NIST)研究了人们对AI的信任问题。
AI的偏见问题会影响人们对其的信任。
人类判断在AI系统得出结论后仍然很重要。
AI系统的发展从描述性到预测性再到规范性。
人们对AI系统的完全信任可能导致问题。
AI系统在特定领域解决问题的能力有限。
人类在AI系统发展中仍扮演着重要角色。
Transcripts
[Music]
how is artificial intelligence being
applied to logistics today my guest is
oliver hedgepeth he is a pro professor
of logistics at the american military
university online hello oliver how are
you
robert i'm doing very well thank you
very much for the invitation to talk
about ai and it is a big deal here and
you are certainly an expert as founding
director of the army's artificial
intelligence center for logistics how do
you see the role of ai in logistics
today
well
started with that job about 1985
and been working
applications of it for ever it seems
like but ai is everywhere today you
don't see it because it's everywhere
we do have it at restaurants we have it
in your home you've got vacuum cleaners
uh in the military and in the civilian
world it's rampant
there are
people people don't realize how much the
military is using ai artificial
intelligence to help soldiers do their
job
and robots as well and surgeons are
using it and restaurateurs are using it
cooks are using it sometimes so it is
everywhere it's expanding
and it's going to continue to expand
it's here with us and there are people
who don't like it or don't like a robot
or don't like smart systems saying
here's what you should do
such as when you type
in a word document on your computer and
you misspell a word and it changes it
for you because it knows you've
misspelled it that's artificial
intelligence people don't realize that
so that's that's interesting
yeah that's such a wide definition
i think maybe the definition of ai has
changed over the years back decades ago
when it was first being developed the
idea was to somehow replicate the way
that the human brain
thought and we were always like 10 years
away from ai becoming a reality and now
all of a sudden it's everywhere but
maybe under slightly
different definition that it's more like
everyday use usefulness now right
well yes it is like everyday use and
you're also correct on the definitions
changed i've got a nice computer
dictionary published in 1967 when i
started in the computer world and it
defined artificial intelligence it
seemed to be pretty good definition
but today it has changed and i did look
up the definition of artificial
intelligence
um a few days ago for a paper i'm
writing and i found six different
definitions it's like really that's kind
of cool because the definitions depend
on who you are robert you have a
different view of what ai might mean to
you i have one for me we're biased in
how we do this there is no one
definition
of ai merriam-webster the dictionaries
do have a definition but there really
isn't just one definition there's
varieties of it and it changes
almost every year now yeah well i'm
interested in what it means to you so
that's kind of the
subject of this conversation today now i
know that ai and robotics are not
completely synonymous but where does
robotics come into the picture when
we're talking about ai
a lot of people don't realize that
robotics is everywhere too and
the military as i mentioned earlier the
military and civilian world the military
for example has
robots that are helping uh fight battles
for them they help them shoot the enemy
we have robotic
drones that can go out and spot an enemy
and shoot them send a missile to them we
have robots in the military field that
when a soldier is walking through high
grass field with his weapon at night and
night cargo is trying to look for the
enemy
well guess what there's a little robot
out there with him
doing the job for him just a little
ahead of him to make sure the man
doesn't walk on a mine or the woman
soldier doesn't walk on a mine and the
robot does it we have robots that will
carry the soldiers
effects you know instead of having to
load all these hundred pound bags on
your back well there are robots to help
do some of that as well so robots are
everywhere yeah
okay civilian world too yeah well let's
talk about the civilian world what are
some of the specific uses in in
logistics
specifically some really exciting and
really far-reaching applications of ai
in that area right now oh gosh that that
is a wonderful field
people think of logistics as
well
ordering something from amazon okay
logistics and amazon works anonymous
amazon uses ai and uses robots to pick
boxes up that they're going to put in a
box and send to you along with the human
that's next to them they use a lot of
robots they use thousands of robots and
it's part of the logistics system they
use ai to track who you are
if you go order something online
your and you pay it by credit card the
software is ai based is instantly
checking everything and making sure you
get what you want and the delivery
schedule comes to you within seconds
where that didn't happen before that's
ai that's doing that
so warehousing is using a lot of it
you go to a grocery store they use ai to
help track
the products that are there in the back
that are on a supply chart coming from
some distance far away or they may be
delayed now that we are living in a
pandemic world this is 2021
there are six months to 12 months delays
on products coming to retail stores to
restaurants
to grocery stores to clothing stores and
so there's ai that's tracking where the
supply chain is because the human has
too much going on there's the supply
chain and the logistics systems are
really too large for one person to have
in their head and to keep sorted out
you've got too much data for too many
thousands there are thousands of supply
chains which have parts that go all the
way back to maybe a little copper mine
in some foreign country that build
digs that copper up that goes and
finally makes it into
that iphone you got yeah and some kind
of oil id oil compound that comes from
maybe alaska oil fields or
iranian oil fields wherever it comes
from or texas
and it's all linked together it's so
it is so complex but it's everywhere and
it's most exciting to see it happen well
it's come it's come a long way in a very
relatively short time but i'm wondering
at this point and of course it's still
evolving
what are the limitations of ai today as
applied to logistics wow the limitations
today uh
are there's a new kind of limitations
happening and it's called i'd say trust
the national institute of standards and
technology it's a department of commerce
company okay you don't think of them as
ai a lot well they just published a
report in june they've got an ai group
that's looking at how people trust
robots how people trust
ai that's helping them do their job live
in their home you know vacuum their
carpet or something or order something
uh a product
they have
studied and started thinking about why
you trust a robot
or why you trust ai system
and they've defined
different aspects of ai you more about
talk about definitions they've got like
30 different categories of
ai
definitions and it's like really strange
to do that but it's trust and it's in it
they've been looking at the trust in
terms of bias see you were biased and i
am biased based on our background if i
build an ai system if you build an ai
system and i build one doing the same
thing
they will not be identical because you
have a different age a different
background you come from california i
came from virginia that may make me do
something with the computer programming
code just a little bit different so that
when it's trying to solve a problem
because ai is solving problems robots
are solving problems it may solve it
just a little bit different for someone
who's using it and that gets toward the
trust factor you may not trust it if
it's giving you an answer that's
slightly different than you think it
should be well
you don't really want to trust it
entirely do you i mean you don't want to
distrust an ai system but isn't there
still a role for humans in terms of what
they do with the conclusion of an ai
system as to how they choose to execute
it or not
so human judgment
yeah
human judgment is still there but i was
looking at a program uh two or three
weeks ago on television of these
musicians i can't remember their name
and they were all excited they had they
put together music and they have all
these instruments and they take weeks
and weeks and weeks to get a new sound
well they bought some kind of ai gadget
that makes sound
and it takes in a matter of seconds to
create this brand new sound effect
a new instrument effect that used to
take two or three weeks and they say ai
is going to solve my problem
they trust it 100 percent so there are
people like yeah i'm turning it over to
you you give me the solution ai machine
you give me the music sound
i want to hear they tell the ai system
here's what i want and they sit back and
they trust it 100 percent now not
everybody does but there are people who
do trust ai 100
but as we go as ai evolves from being a
descriptive system
to a predictive system and finally to a
prescriptive system where it's actually
telling you what actions to take
do we want to leave that up to the
machine and if we do want to leave it up
the machine do we want to know how it
arrived at that conclusion or are we
comfortable with this idea of the black
box just outputting a conclusion that we
should go ahead and implement because we
know they're smarter than we are
i i'm glad you asked that question uh
and that was part of the uh report from
the national institute of standards of
technology published just last june
from their ai task force in which
you don't have to worry about that um
there was a comment made by some one of
the authors about it may be 20 or 30 or
40 50 years from now before you get to
that stage the ai systems are still
they're strictly prescriptive they go
from here to here you give them a
problem to solve and they solve that
problem you know they may be able to
play chess with you and beat the heck
out of you
but they can't clean the carpet
okay or they can't recommend uh an
ingredient for
spaghetti and meatballs they might they
can't you know they may not be able to
do that they can't do everything it's
not like a human brain in the ai system
or even a robot that does something
they're limited
they're very prescriptive it's almost
like medicine they're very prescriptive
to solve a problem you got a headache
you take an aspirin you got a problem to
solve that's really unique for this
robot or this ai system you turn it on
and it gives you a solution
but
you as the human have to sit back and
say
do i really want to do that
uh i'll give you one example
there was a robot that was
being used to guide people out of
burning buildings it's dark and smoky
and people not firemen
but people are in there and it's smoky
and it's like a big building in downtown
new york or somewhere would be on fire
and you want a robot to guide you
through this dark
way to get out the robot was guiding the
people into
closed rooms
you know it didn't work all the time
ninety percent of the time it got them
out the door you know it got into it
like this is not a situation where ten
percent is an acceptable margin they're
like
they got it they read the robot read
something wrong and whatever or read
something differently maybe i won't say
wrong read it differently and it went
into a closet with a bunch of people now
here's the key thing
those people who got into the closet
they stayed there
thinking well the robot's gonna fix it
now i'm gonna wait here he's gonna tell
us to go out and go to the right place
any minute now we won't go looking
wow
okay
well the message here is that ai has
come a long way it does great things but
still a role for humans to play as ai
even as ai develops uh dr oliver
hedgepeth uh i really appreciate whether
the american military university online
i really appreciate your participation
in this interview and your insights into
this into this issue thank you so much
for being with me today well thank you
very much for the questions
you
Weitere ähnliche Videos ansehen
In conversation with the Godfather of AI
2 Ex-AI CEOs Debate the Future of AI w/ Emad Mostaque & Nat Friedman | EP #98
AI and Quantum Computing: Glimpsing the Near Future
AGI: solved already?
Geoffrey Hinton in conversation with Fei-Fei Li — Responsible AI development
【人工智能】2027年将会实现AGI | 约翰·舒尔曼最新访谈 | OpenAI联合创始人&首席架构师 | 大模型训练 | RLHF | 技术瓶颈 | 模型蒸馏 | AI监管
5.0 / 5 (0 votes)