| of our artificial intelligence the first |
5.641 |
Jump to 530.279 |
ORDINAL
first
|
| like we do the first cell phone those |
4.86 |
Jump to 678 |
ORDINAL
first
|
| for example section 230 forcing |
5.159 |
Jump to 969.12 |
CARDINAL
230
|
| one is only the first |
4.38 |
Jump to 1040.819 |
ORDINAL
first
|
| looking at one of the most significant |
4.98 |
Jump to 1116.66 |
CARDINAL
one
|
| the story of the 20th century largely |
3.659 |
Jump to 1207.26 |
DATE
the 20th century
|
| attention for almost two decades and |
4.681 |
Jump to 1270.679 |
DATE
almost two decades
|
| we had four bills initially that were |
5.94 |
Jump to 1277.58 |
CARDINAL
four
|
| all four bills with unanimous roll calls |
5.579 |
Jump to 1286.34 |
CARDINAL
four
|
| 1989 was a historic Watershed year in |
6.219 |
Jump to 1306.82 |
DATE
1989
DATE
Watershed year
|
| in 1989 and I think about those in |
4.679 |
Jump to 1326.12 |
DATE
1989
|
| point and the second starting point is |
3.901 |
Jump to 1375.679 |
ORDINAL
second
|
| one that's humbling and that is effect |
3.959 |
Jump to 1378.08 |
|
| accelerator why combinator from 1914 I'm |
8.579 |
Jump to 1487.1 |
DATE
1914
|
| sorry 2014 to 2019 open AI was founded |
7.141 |
Jump to 1491.059 |
DATE
2014 to 2019
GPE
AI
|
| in 2015. |
4.62 |
Jump to 1495.679 |
DATE
2015
|
| committee which was established in 2022 |
5.28 |
Jump to 1540.2 |
DATE
2022
|
| we are working to build tools that one |
3.78 |
Jump to 1674.72 |
CARDINAL
one
|
| we spent over six months conducting |
4.92 |
Jump to 1750.98 |
DATE
six months
|
| such an approach would involve four |
3.601 |
Jump to 1940.399 |
CARDINAL
four
|
| first different rules for different |
2.88 |
Jump to 1944 |
ORDINAL
first
|
| second clearly defining risks there must |
4.92 |
Jump to 1954.32 |
ORDINAL
second
|
| third be transparent so AI shouldn't be |
4.441 |
Jump to 1971.899 |
ORDINAL
third
GPE
AI
|
| anybody can deny anything as one |
3.72 |
Jump to 2187.42 |
CARDINAL
one
|
| person posing as a 13 year old how to |
4.5 |
Jump to 2228.46 |
DATE
13 year old
|
| 31 year old man |
3.72 |
Jump to 2232.96 |
DATE
31 year old
|
| need to generate Financial return seven |
4.08 |
Jump to 2314.74 |
CARDINAL
seven
|
| evaluation is a vital First Step |
4.98 |
Jump to 2368.68 |
ORDINAL
First
|
| Marcus we're going to have seven minute |
5.18 |
Jump to 2418.96 |
TIME
seven minute
|
| first of all Professor Marcus we're here |
5.86 |
Jump to 2424.14 |
ORDINAL
first
PERSON
Marcus
|
| first of all I think it's important to |
3.601 |
Jump to 2729.819 |
ORDINAL
first
|
| they use it and second gpt4 and things |
6.899 |
Jump to 2740.92 |
ORDINAL
second
|
| technology is one long technological |
3.72 |
Jump to 2769.96 |
CARDINAL
one
|
| hugely important question and it's one |
4.859 |
Jump to 2820.48 |
CARDINAL
one
|
| we created you know as early as three |
4.26 |
Jump to 2849.04 |
|
| seven million Learners and over a |
4.079 |
Jump to 2881.74 |
CARDINAL
seven million
|
| thousand courses worldwide focused on |
5.82 |
Jump to 2883.48 |
|
| skills and we've pledged to train 30 |
6.661 |
Jump to 2885.819 |
CARDINAL
30
|
| million individuals by 2030 in the |
5.46 |
Jump to 2889.3 |
CARDINAL
2030
|
| to the first question as well absolutely |
4.921 |
Jump to 2897.339 |
ORDINAL
first
|
| first of all that scientists be part of |
3.78 |
Jump to 2924.7 |
ORDINAL
first
|
| that process and second that we have |
3.54 |
Jump to 2926.38 |
ORDINAL
second
|
| second is on jobs past performance |
5.26 |
Jump to 2943.2 |
ORDINAL
second
|
| technologies come in I think this one's |
3.179 |
Jump to 2955.66 |
|
| going to be 10 years is it going to be |
2.639 |
Jump to 2961.18 |
DATE
10 years
|
| 100 years and I don't think anybody |
3.3 |
Jump to 2962.44 |
DATE
100 years
|
| 20 years people will laugh at this as I |
4.38 |
Jump to 2982.3 |
DATE
20 years
|
| look back at the AI of today 20 years |
4.98 |
Jump to 2991.359 |
FAC
the AI of
DATE
20 years
|
| intelligence maybe let's say it's 50 |
3.239 |
Jump to 3007.38 |
CARDINAL
50
|
| a month ago the authors are two Andreas |
7.201 |
Jump to 3164.099 |
DATE
a month ago
CARDINAL
two
GPE
Andreas
|
| three years ago now about the effect of |
5.581 |
Jump to 3242.099 |
DATE
three years ago
|
| Senator Hawley for the question it's one |
4.02 |
Jump to 3292.38 |
PERSON
Hawley
CARDINAL
one
|
| sort of one-on-one uh you know |
3.66 |
Jump to 3301.74 |
CARDINAL
one
|
| address this yeah I'd like to add two |
3.421 |
Jump to 3396.359 |
CARDINAL
two
|
| things one is in the appendix to my |
4.019 |
Jump to 3397.8 |
|
| remarks I have two papers to make you |
4.02 |
Jump to 3399.78 |
CARDINAL
two
|
| um one is in the Wall Street Journal |
4.019 |
Jump to 3403.8 |
ORG
the Wall Street Journal
|
| ultimately to data one of the things |
4.621 |
Jump to 3437.579 |
|
| each one of us |
4.62 |
Jump to 3527.28 |
|
| space isn't one that we necessarily |
4.38 |
Jump to 3665.04 |
|
| as I mentioned section 230 in my opening |
6.059 |
Jump to 3753.78 |
CARDINAL
230
|
| we decided that in section 230 that we |
5.581 |
Jump to 3759.839 |
LAW
section 230
|
| that section 230 doesn't apply to |
4.92 |
Jump to 3777.54 |
LAW
section 230
|
| so what have we learned from 230 that |
6.659 |
Jump to 3788.7 |
CARDINAL
230
|
| there is one model here where we stick |
5.339 |
Jump to 4031.28 |
|
| one at least in the United States but |
4.439 |
Jump to 4078.68 |
CARDINAL
one
GPE
the United States
|
| train and you have to have 190 some of |
5.16 |
Jump to 4144.44 |
CARDINAL
190
|
| them you know one for every country that |
4.259 |
Jump to 4146.96 |
|
| things first but to be effective uh we |
4.98 |
Jump to 4205.1 |
|
| it gave me four Pros four cons and says |
5.401 |
Jump to 4287.179 |
CARDINAL
four
CARDINAL
four
|
| talked with them Mr chairman one of the |
5.46 |
Jump to 4354.02 |
CARDINAL
one
|
| have a second round I've got a host of |
5.82 |
Jump to 4410.06 |
ORDINAL
second
|
| I would say first of all we think that |
5.041 |
Jump to 4516.739 |
ORDINAL
first
|
| um second I think that we need to figure |
4.56 |
Jump to 4529.88 |
|
| that would be a third party entity okay |
6.96 |
Jump to 4570.08 |
ORDINAL
third
|
| without first receiving their consent so |
5.879 |
Jump to 4596.3 |
ORDINAL
first
|
| first of all jukebox is not a product we |
3.54 |
Jump to 4600.38 |
ORDINAL
first
|
| this uh number one we think that people |
4.2 |
Jump to 4751.4 |
|
| artists of all time and two of the top |
4.859 |
Jump to 4786.86 |
CARDINAL
two
|
| three were from Minnesota that would be |
5.34 |
Jump to 4789.02 |
CARDINAL
three
GPE
Minnesota
|
| us continue one thing AI won't change |
5.4 |
Jump to 4797.12 |
ORG
AI
|
| Grahams of section 230 is that we just |
5.46 |
Jump to 4830.12 |
LAW
section 230
|
| to focus in on what my job one of my |
3.66 |
Jump to 4837.62 |
|
| fake thing go to one two three four Elm |
6.42 |
Jump to 4864.56 |
CARDINAL
one two
CARDINAL
four
ORG
Elm
|
| one tweet is okay yeah and of course |
3.48 |
Jump to 4964.94 |
CARDINAL
one
|
| those fake ads so that's number one |
6.3 |
Jump to 4980.3 |
CARDINAL
number one
|
| number two is the impact on intellectual |
6.48 |
Jump to 4982.34 |
CARDINAL
number two
|
| one-third of the U.S newspapers are that |
5.04 |
Jump to 5027.82 |
CARDINAL
one-third
GPE
U.S
|
| roughly existed two decades are going to |
5.64 |
Jump to 5030.34 |
DATE
two decades
|
| go are going to be gone by 2025 unless |
5.58 |
Jump to 5032.86 |
DATE
2025
|
| section 230 but I think asking little |
5.76 |
Jump to 5052.3 |
LAW
section 230
|
| round one of the internet has not been |
3.239 |
Jump to 5075.46 |
CARDINAL
one
|
| gpt4 ended training in 2021 it's not |
4.8 |
Jump to 5107.26 |
DATE
2021
|
| office showing that something like 50 |
3.92 |
Jump to 5173.98 |
|
| basically section 230 is being used by |
5.941 |
Jump to 5232.179 |
LAW
section 230
|
| and she calls three times she gets no |
5.76 |
Jump to 5256.3 |
CARDINAL
three
|
| if I may speak for one second there's a |
3.619 |
Jump to 5271.54 |
TIME
one second
|
| section 230 applies to the tool you have |
4.679 |
Jump to 5299.02 |
LAW
section 230
|
| I don't think section 230 is the even |
6.301 |
Jump to 5307.179 |
LAW
section 230
|
| like section 2 the 230 applies to your |
4.799 |
Jump to 5360.4 |
LAW
section 2
CARDINAL
230
|
| three time tested ways to protect |
4.381 |
Jump to 5370.659 |
CARDINAL
three
|
| but nuclear power could be one of the |
5.399 |
Jump to 5396.28 |
|
| the technology so should we have one but |
7.079 |
Jump to 5455.08 |
|
| to produce one of these tools I think it |
4.98 |
Jump to 5476.62 |
CARDINAL
one
|
| you claim section 230 applies in this |
4.86 |
Jump to 5488.02 |
CARDINAL
230
|
| the domains and the issues this one with |
5.1 |
Jump to 5524.32 |
|
| so here's what my two cents worth for |
4.62 |
Jump to 5561.34 |
MONEY
two cents
|
| 15 military application |
7.32 |
Jump to 5611.26 |
CARDINAL
15
|
| and you got one minute |
3.78 |
Jump to 5618.58 |
TIME
one minute
|
| I got one minute yeah all right this is |
4.681 |
Jump to 5620.199 |
TIME
one minute
|
| that's a tough question for one minute |
4.799 |
Jump to 5622.36 |
TIME
one minute
|
| expertise uh but I give you one example |
5.281 |
Jump to 5627.159 |
CARDINAL
one
|
| wild I understand one way to prevent |
4.5 |
Jump to 5776.32 |
CARDINAL
one
|
| spent well over six months with gpt4 |
4.261 |
Jump to 5882.219 |
DATE
well over six months
|
| address those one of the things that's |
2.94 |
Jump to 5896.5 |
|
| influence elections as one of the |
4.079 |
Jump to 6003.6 |
CARDINAL
one
|
| highest risk cases one that's entirely |
4.86 |
Jump to 6005.4 |
|
| direction is the right one to pursue |
5.94 |
Jump to 6023.1 |
|
| conversation and one that could also |
4.621 |
Jump to 6150.179 |
|
| permit me to share with you three |
6 |
Jump to 6221.76 |
CARDINAL
three
|
| hypothesis number one |
5.561 |
Jump to 6231.739 |
CARDINAL
number one
|
| hypothesis number two |
7.66 |
Jump to 6242.84 |
CARDINAL
number two
|
| hypothesis number three that I would |
5.5 |
Jump to 6266.78 |
CARDINAL
number three
|
| two or three reforms regulations if any |
8.04 |
Jump to 6303.54 |
CARDINAL
two
CARDINAL
three
|
| these okay so we ought to first pass the |
4.259 |
Jump to 6382.86 |
ORDINAL
first
|
| in the first place as well Professor |
4.741 |
Jump to 6407.639 |
ORDINAL
first
|
| looking for specificity number one a |
6.12 |
Jump to 6424.38 |
CARDINAL
number one
|
| going to introduce something to 100 |
4.14 |
Jump to 6432.84 |
CARDINAL
100
|
| million people somebody has to have |
4.56 |
Jump to 6434.94 |
|
| that's a good one number sure I agree |
4.26 |
Jump to 6439.5 |
|
| with it but that's a good one what else |
3.24 |
Jump to 6441.84 |
|
| you didn't ask for three that you would |
3.479 |
Jump to 6443.76 |
CARDINAL
three
|
| agree with number two |
5.039 |
Jump to 6445.08 |
CARDINAL
number two
|
| three would be funding geared towards |
5.46 |
Jump to 6459.06 |
CARDINAL
three
|
| Senator uh number one I would form a new |
4.2 |
Jump to 6497.88 |
CARDINAL
number one
|
| two I would create a set of safety |
3.601 |
Jump to 6507.239 |
CARDINAL
two
|
| your third hypothesis as the dangerous |
4.379 |
Jump to 6510.84 |
ORDINAL
third
|
| capability evaluations one example that |
3.96 |
Jump to 6512.699 |
CARDINAL
one
|
| into the world and then third I would |
4.681 |
Jump to 6527.159 |
ORDINAL
third
|
| it might have been I don't know one of |
2.781 |
Jump to 6610.139 |
|
| that can be done and one of the things |
4.26 |
Jump to 6656.52 |
CARDINAL
one
|
| I'll give a few examples uh one would be |
4.38 |
Jump to 6679.02 |
|
| one of us that could be put out there |
4.141 |
Jump to 6714.119 |
CARDINAL
one
|
| you have it so two of two of you said |
5.52 |
Jump to 6720.54 |
CARDINAL
two
CARDINAL
two
|
| the two of you even Envision as a |
4.44 |
Jump to 6748.38 |
CARDINAL
two
|
| the first part of what what you said |
4.44 |
Jump to 6761.219 |
ORDINAL
first
|
| contemplate well first if I can rewind |
4.62 |
Jump to 6844.8 |
ORDINAL
first
|
| important on this second part of your |
4.799 |
Jump to 6878.94 |
ORDINAL
second
|
| one specific example Auto GPT frightens |
4.44 |
Jump to 6909.48 |
CARDINAL
one
PRODUCT
Auto GPT
|
| of AI so we were one of the first teams |
6.301 |
Jump to 7221.599 |
GPE
AI
ORDINAL
first
|
| can you speak just for a second |
4.861 |
Jump to 7238.219 |
ORDINAL
second
|
| this is really important one example is |
3.601 |
Jump to 7264.679 |
CARDINAL
one
|
| want to ask one more question this |
4.8 |
Jump to 7401.96 |
CARDINAL
one
|
| uh the Europe has has rolled one out to |
8.16 |
Jump to 7721.82 |
LOC
Europe
CARDINAL
one
|
| mixed reviews do you think we need one I |
5.621 |
Jump to 7726.699 |
|
| certainly similar tools one of the |
5.101 |
Jump to 7775.219 |
|
| other words one cannot |
2.879 |
Jump to 7852.26 |
|
| to be 18 or up or have your parents |
3.6 |
Jump to 7878.96 |
CARDINAL
18
|
| permission at 13 and up to use a product |
3.6 |
Jump to 7880.82 |
CARDINAL
13
|
| I first of all I think we try to design |
5.339 |
Jump to 7957.86 |
ORDINAL
first
|
| second uh |
4.86 |
Jump to 7977.48 |
ORDINAL
second
|
| all of us one of the things that we'll |
4.379 |
Jump to 7990.86 |
|
| has been one of the best hearings I've |
2.94 |
Jump to 8036.219 |
|
| the child of two IBM parents |
4.88 |
Jump to 8132.4 |
CARDINAL
two
ORG
IBM
|
| we can try to work on two different ways |
6.54 |
Jump to 8159.4 |
CARDINAL
two
|
| one thing I haven't as much as I would |
3.42 |
Jump to 8277.599 |
CARDINAL
one
|
| um first of all you're a bit of a |
2.939 |
Jump to 8313.78 |
ORDINAL
first
|
| unicorn when I said that with you first |
4.441 |
Jump to 8315.219 |
ORDINAL
first
|
| the second part of my question as well |
5.941 |
Jump to 8364.719 |
ORDINAL
second
|
| one of my biggest concerns about this |
4.18 |
Jump to 8393.84 |
CARDINAL
one
|
| space of uh web 2 web 3 is this massive |
5.64 |
Jump to 8398.02 |
CARDINAL
2
CARDINAL
3
|
| that scares me Sam I'm sorry one more |
4.199 |
Jump to 8492.7 |
PERSON
Sam
CARDINAL
one
|
| thing I wanted to add one thing that I |
3.54 |
Jump to 8495.1 |
CARDINAL
one
|
| first of all I want to thank you Senator |
5.2 |
Jump to 8525.06 |
ORDINAL
first
|
| minute of it you have one of our longer |
5.94 |
Jump to 8539.08 |
|
| and number one you're here because AI is |
7.381 |
Jump to 8563.319 |
CARDINAL
number one
ORG
AI
|
| number two is really unknown what's |
4.56 |
Jump to 8576.64 |
CARDINAL
number two
|
| uh number three |
5.439 |
Jump to 8590.46 |
CARDINAL
number three
|
| and the two things that I want to ask |
4.92 |
Jump to 8674.74 |
CARDINAL
two
|
| one you've somewhat answered because I |
4.081 |
Jump to 8677.439 |
|
| think two of the three of you have said |
3.659 |
Jump to 8679.66 |
CARDINAL
two
CARDINAL
three
|
| two of the three of you have said you |
4.26 |
Jump to 8736.78 |
CARDINAL
two
CARDINAL
three
|
| one I think America has got to continue |
5.52 |
Jump to 8763.54 |
GPE
America
|
| that's one I mean that is apparel which |
3.18 |
Jump to 8806.5 |
|
| a second and I think this can happen |
3.66 |
Jump to 8813.34 |
ORDINAL
second
|
| us could handle but a smaller one |
3.721 |
Jump to 8843.939 |
CARDINAL
one
|
| one of the things I would add to the |
3.54 |
Jump to 8877 |
CARDINAL
one
|
| at all but one was uh just alluded to |
8.579 |
Jump to 8945.28 |
|
| 20 years I was a federal prosecutor the |
4.62 |
Jump to 9059.88 |
DATE
20 years
|
| create 10 new agencies |
4.979 |
Jump to 9067.38 |
CARDINAL
10
|
| regulation you can think of five |
3.06 |
Jump to 9098.28 |
CARDINAL
five
|
| first Do no harm |
5.94 |
Jump to 9137.7 |
ORDINAL
first
|
| you know literally 15 years ago I think |
5.34 |
Jump to 9169.979 |
DATE
15 years ago
|
| abdicated abolishing section 230. |
5.52 |
Jump to 9172.26 |
LAW
section 230
|
| abolishing section 230 back then it was |
6.88 |
Jump to 9180.8 |
LAW
section 230
|
| uh one is that we don't train on any |
4.98 |
Jump to 9211.56 |
|
| retain it for 30 days solely for the |
4.62 |
Jump to 9220.08 |
DATE
30 days
|
| one I think we don't really understand |
3.54 |
Jump to 9286.26 |
|
| one could imagine a lot more and that |
3.361 |
Jump to 9295.439 |
|
| could happen in two years it could |
4.981 |
Jump to 9297.479 |
DATE
two years
|
| happen in 20 years the basic paradigms |
5.519 |
Jump to 9298.8 |
DATE
20 years
|
| one second one thing that is absolutely |
4.02 |
Jump to 9310.26 |
QUANTITY
one second
|
| what's in one of these systems but I |
3.179 |
Jump to 9321.18 |
|
| is hugely important one and coming back |
5.04 |
Jump to 9366.06 |
|
| concern in my opening I suggested three |
6.061 |
Jump to 9499.439 |
CARDINAL
three
|
| is Montgomery 100 and as you also |
5.52 |
Jump to 9517.2 |
ORG
Montgomery 100
|
| think those three would be a great start |
3.12 |
Jump to 9529.8 |
CARDINAL
three
|
| that it's it's potentially laying off 7 |
4.319 |
Jump to 9638.04 |
CARDINAL
7
|
| 800 people a third of your non-consumer |
5.579 |
Jump to 9639.78 |
CARDINAL
800
|
| about a thousand technology and AI |
5.339 |
Jump to 9664.14 |
CARDINAL
about a thousand
ORG
AI
|
| six-month moratorium on any further AI |
5.28 |
Jump to 9671.46 |
DATE
six-month
ORG
AI
|
| should we pause for six months |
4.08 |
Jump to 9679.8 |
DATE
six months
|
| I actually signed that letter about 27 |
3.96 |
Jump to 9683.88 |
CARDINAL
about 27
|
| 000 people signed it |
4.86 |
Jump to 9686.16 |
CARDINAL
000
|
| moratorium a six-month moratorium or |
5.639 |
Jump to 9714.72 |
DATE
six-month
|
| deployed at a scale of let's say 100 |
3.719 |
Jump to 9753.6 |
CARDINAL
100
|
| million people without any external |
4.319 |
Jump to 9755.04 |
|
| further development for six months or |
4.14 |
Jump to 9764.52 |
DATE
six months
|
| longer uh so first of all we after we |
4.62 |
Jump to 9766.08 |
ORDINAL
first
|
| six months to deploy it |
3.18 |
Jump to 9770.7 |
DATE
six months
|
| it in the next six months but I think |
4.32 |
Jump to 9775.68 |
DATE
the next six months
|
| we pause for six months then I'm not |
3 |
Jump to 9787.2 |
DATE
six months
|
| pause for another six do we kind of come |
4.56 |
Jump to 9790.2 |
CARDINAL
six
|
| new one for deploying as I mentioned I |
2.7 |
Jump to 9823.5 |
|
| history for 100 years maybe this agency |
2.88 |
Jump to 9882.54 |
DATE
100 years
|
| by section 230 but there's not currently |
4.98 |
Jump to 9933.6 |
LAW
section 230
|
| know that 230 does or does not apply |
4.019 |
Jump to 10030.92 |
CARDINAL
230
|
| easy for us to say that section 230 |
3.42 |
Jump to 10046.16 |
CARDINAL
230
|
| things at a scale of a million people or |
3.54 |
Jump to 10134.96 |
CARDINAL
million
|
| 10 million people but not a hundred |
3.481 |
Jump to 10136.819 |
CARDINAL
10 million
CARDINAL
a hundred
|
| million people or a billion people and |
3.3 |
Jump to 10138.5 |
|
| the last 25 years is is what really |
4.439 |
Jump to 10211.46 |
DATE
the last 25 years
|
| concerns me and one of the things |
5.28 |
Jump to 10213.68 |
|
| New Media a 24-hour cable news is a |
5.279 |
Jump to 10223.16 |
TIME
24-hour
|
| way I thought it would in one one way it |
3.66 |
Jump to 10282.08 |
CARDINAL
one
|
| record open for one week in case anyone |
4.379 |
Jump to 10703.1 |
DATE
one week
|
| forward to our next hearing this one is |
4.7 |
Jump to 10719.18 |
|