Author of "The Algorithm" - Hilke Schellmann - Career Club Live PART 1

[00:00:00] Hello everybody this is Bob Goodwin and welcome to another episode of Career Club Live.

[00:00:15] Before we begin we've got a couple new things to tell you about a career club.

[00:00:19] If you're a job seeker we are now offering some free resources including a free weekly

[00:00:24] coaching call every Thursday at 1 o'clock Eastern just go to career.club, click on for job

[00:00:30] seekers free resources and you'll see it and if you're a company we're starting something new

[00:00:35] to add to the candidate experience and help build your employer brand by changing

[00:00:40] rejection letters into read direction letters providing free career resources to people who

[00:00:45] will not be moving forward in the hiring process. It's a great way to build your brand and have a

[00:00:50] differentiated candidate experience again you can learn more about that at career.club,

[00:00:55] under four employers and then under candidate experience. So with that said commercial

[00:01:02] over I am extremely excited to welcome our guests today, Hilke Schelman. I'm a real bit about this because

[00:01:08] her background is phenomenal. So Hilke is in any award-winning journalist in the insightful author of

[00:01:15] the recently published book The Algorithm how the rise of artificial intelligence will change

[00:01:20] everything and what we can do to protect our future. Her groundbreaking work has been featured in

[00:01:25] prestigious outlets like The Wall Street Journal, Time Magazine, The New York Times, The Guardian,

[00:01:30] The MIT Technology Review and it delves into the profound impact of artificial intelligence

[00:01:36] on the workforce and society. Besides her writing, Hilke is also a professor at NYU where she

[00:01:41] shapes the minds of future journalists with her deep understanding of AI's ethical implications

[00:01:46] in transformative potential. So today we're going to explore the depths of AI's influence on

[00:01:50] our world navigating the complex interplay between technology and human rights. With that,

[00:01:56] let me welcome Hilke. Oh wonderful. Thank you so much for that kind of introduction Bob. I'm so delighted

[00:02:03] to be here. No likewise. And like I love all my guests. I am particularly over the moon today because

[00:02:10] in all seriousness, you know, Hilke your topic. AI obviously that's dominating headlines.

[00:02:16] But our core audience is both job seekers and HR executives. And so this is like the perfect

[00:02:23] intersection of all of those things and you are a global expert on the topic. So it's going to be

[00:02:30] super, super fun. But before we dive into all that as we are want to do, would love to just

[00:02:35] sort of a little bit more about you as a human being. So just what I'm not in so much.

[00:02:41] Exactly. Where do we find you today? Where are you? I'm in Brooklyn, New York.

[00:02:45] I'm in my home office. I was four year old as I'm from Daker. So we all home, don't spring break.

[00:02:54] Yes, all the fun. I'm sure it feels like spring break up. But okay. Now, are you born in

[00:03:00] born and raised New Yorker? No. I'm born and raised in Germany. I went to high school all the way up

[00:03:05] to high school there. Got my bachelor's and master's degree at the Humboldt University in Berlin.

[00:03:11] But I got a full-buy to go to NYU in 2003 and that started all of this, you know, great

[00:03:18] American adventure for me. I started a documentary art collaborative and then I went back to Berlin

[00:03:24] and studied investigative journalism at Columbia, St. Skoleness somehow got trapped into

[00:03:29] staying. I don't know how you figured out the accent, but it's spot-on. It is absolutely spot-on.

[00:03:35] How I will tell my friends who make fun of my German accent. You don't have one.

[00:03:39] So, you anticipate one of the questions? Tell me a little bit about your family.

[00:03:42] So the four year old? Yes. She just turned four. She's four going on 15. She thinks she

[00:03:48] she's ready to move out and conquer the world. We are just like, you know, a little help us on the

[00:03:52] way. She's got to be like, so yeah, I live with a partner. I have a, like, you daughter who was

[00:03:58] born three weeks before the lockdown and the pandemic started. So she's like, you know, we

[00:04:04] will always be with my every birthday and like, oh yeah, that, that happened. Yeah, and so I,

[00:04:10] you know, I work as a journalist most days of the week and I teach usually two classes at

[00:04:16] NYU and sort of divide my time in like the best possible way. I get to do awesome journalism

[00:04:21] and I get to teach journalists. So it's, you know, I found I found my dream job.

[00:04:26] Well, it's just awesome. And we love people finding their dream jobs at career club.

[00:04:31] I mentioned the Emmy Award. I mean, that's not something that you just get to kind of say all the

[00:04:35] time. Do you mind telling people just a little bit about whether or not? Yeah, totally. I'd be, I'd be,

[00:04:39] I'd be delighted to. So I think, you know, in my previous life as a journalist, I was always a journalist

[00:04:45] but I did a lot of documentary work and one of the documentaries we did was it was my

[00:04:52] co-director and I went into Pakistan and followed a case on sexual assault and in sexual violence

[00:04:58] and in Pakistan and sort of film for like three, three years. This girl story going through

[00:05:03] the legal system and all of the pushback you see from society and especially men. And we did a

[00:05:11] film out of that and, you know, we got lucky that Sundance showed it at the film festival

[00:05:17] and frontline jump from the chance to to broadcast it on a PBS and, you know, we got an Emmy

[00:05:25] for that. So that was really, really fantastic. That is very impressive and obviously you did

[00:05:32] and rift on your laurels you moved on to a new topic of interest to you, which is what we're going

[00:05:36] to talk about. Yeah, yeah, you know, sort of founded by my chance a little bit, right? Like I've

[00:05:42] been doing, I've been doing like more or less social justice documentaries and podcast and then one

[00:05:47] day in 2017 November 2017 I was at a conference in Washington DC talking to consumer lawyers

[00:05:53] had no idea about AI and you know, nothing to do with hiring. And I needed a train ride

[00:05:59] from the conference to the train station and I talked to the lift driver and I just, you know,

[00:06:03] asking like how was your day? And he was like, you know, my day is really weird. And I was like,

[00:06:07] what, you know, how so? And he said that he was in a view by a robot for a job as a baggage hand

[00:06:14] and has like what robot doing shopping if you've never heard of that. So I think, you know, what it was

[00:06:19] is like sort of a pre-recorded voice asking him three questions. But, you know, I was really

[00:06:26] interested in that and then I went down the rabbit hole but, you know, and learn how, like, how much we

[00:06:32] use AI and hiring and it work and, you know, now so we see signals of AI, of his else being

[00:06:38] being used in fire rings. So that was sort of like the beginning that I was like, wow,

[00:06:43] a met someone who did a job interview with the Konakura robot and I had never heard of that

[00:06:48] and loan behold there's a whole world out there. Side note for listeners, this is like a really

[00:06:54] interesting example to me if not working. Right, you meet people, you can get some

[00:06:58] conversation that you never know the way it's going to break. It's always kind of this

[00:07:02] weird fractal like I wouldn't expect them out but that's cool and but you have to put yourself

[00:07:07] out there ask questions be interested and you never know what you're going to do. Yeah, you

[00:07:12] never know who knew right? Like, you know, they talked about, you know, and then I went to a

[00:07:16] conference on AI and, you know, the topic came up again and as sparsely attended presentation and I was

[00:07:22] like, oh somebody else was talking about some of the things this now somebody who had just left

[00:07:27] the Equal Employment of Petunay the Commission and she said that she can't sleep at night because

[00:07:31] companies use like basic algorithms to check for absenteeism in people's calendars of their workers

[00:07:37] and she was really worried that this might hurt mothers and people with disabilities who often

[00:07:42] have, you know, longer absentee weights than others but they're protected classes. You're not allowed

[00:07:47] to fire them or do anything against them if there is more time than others so she was really

[00:07:54] worried about that. That's you should be. So what's one to take back here a little bit because I

[00:07:59] want to, you know, dig into it but just to kind of completely level set because the book just

[00:08:06] came out at the beginning of January right? Yeah, January second is like the new year here it is.

[00:08:11] So I mean, the first question I don't mean for to be completely obvious but why this book and why now?

[00:08:18] You know, I've been working on this for like five years or so now, right? Like researching this

[00:08:23] but I felt like there's really like the world is really changing here and we now have all these

[00:08:29] indications that almost all Fortune 500 companies use these kinds of AI hiring tools somewhere

[00:08:35] in their hiring funnel right from like resume screening to even finding folks to reach out to

[00:08:42] to like ask them to do pre one way video interviews right where there's no human on the other side

[00:08:48] to play games. So really felt like a vehicle sea change and also we see that like for workplace

[00:08:54] monitoring and surveillance, we see eight out of the ten largest companies in the US do

[00:09:00] monitor some of the workers and we now see that like really really large companies use some

[00:09:05] of these AI vendors to check for sentiment analysis and check all of our writing. So it feels like

[00:09:10] you know, like AI has taken over the world of work and how into some of these AI job conference

[00:09:15] and it's like all these hundreds of vendors are all AI and I think you know when I talked to

[00:09:20] job applicants I was surprised that they sort of just didn't know that this has happened that

[00:09:26] you know somewhere maybe vaguely a new that maybe if they post something on LinkedIn,

[00:09:32] you know there there was a maybe screen by AI but I think no no one knew like how vast this

[00:09:37] industry has mushrooms and we didn't see a lot of that. So you know, when you do these conferences

[00:09:44] there's like thousand, ten thousand people go to HR tech every year and let's make it.

[00:09:48] And I was like wow there's like a couple of reporters there isn't a whole lot going on here but

[00:09:53] it feels like we are changing so much we're trying to quantify human beings in a mathematical way

[00:10:01] much more so than we ever could you know obviously we you know the fields of psychometric

[00:10:05] success that's tried that for a long time but I felt like this is a real sea change and we need

[00:10:10] to we need to talk about it now. No I think that's right and so again we're going to get into a

[00:10:14] lot of this more. So you know as as you think about how AI has been used and what's just sort

[00:10:21] of start if the start you know somebody does a job posting they've right they've got their

[00:10:27] applicant tracking system in place. I think most listeners would know the term ATS so there's an

[00:10:35] ATS and its job is to read the flood of resumes coming in and you at least start to kind

[00:10:43] of have an in and an out pile right like yes yes you're going to buy for this job this is a brain

[00:10:49] surgeon and you're a bricklayer probably not a good fit and in theory you know, shunt the right

[00:10:55] ones over to you know town acquisition or hiring manager or whatever that's the theory how's it work

[00:11:05] I mean you know we we don't know a whole lot about this world right like we know I talk to all

[00:11:11] the large job platforms like LinkedIn and indeed and all of them use some form of AI so if you submit your

[00:11:19] resume via them your resume will not likely hold check by and we know that companies use these

[00:11:28] applicant tracking systems and I think they used to be sort of glorified excel sheets so just

[00:11:33] checked like where is a candidate in the hiring funnel but now we know from all of the largest

[00:11:38] vendors that they have AI built in but you know there's no central place where companies need to

[00:11:43] tell anyone the government or anyone what they're doing so we don't actually know

[00:11:48] what do the companies turn on and off which I do other using how's it being trained you know

[00:11:53] what's the training data which is really important in in resume screening or in AI tools in

[00:12:00] general so what I know is I talk to a bunch of employment lawyers and others who sort of

[00:12:06] they're at the moment when maybe a vendor you know has a resume screening tool and they want

[00:12:11] a shelter company and a company is doing some due diligence that may bring in employment lawyers

[00:12:16] a folks like in in that space to do their due diligence and what they have found is like

[00:12:23] a little bit worrisome sometimes so one of the person I talked to Dr. John Scott he said that

[00:12:31] he found he checked five resume screeners and five all of them had had had problems in that

[00:12:37] bias variable so one of them had learned learned that the first name Thomas was a predictive success

[00:12:46] another one had that like Syria and the word Syria and the word Canada on a resume was an indicator

[00:12:54] of success so you know when I talked to employment lawyers they were like this could be

[00:12:58] discrimination based on national origin like it's not allowed to take locations into consideration

[00:13:06] you know we saw another employment lawyer found the word Africa and African American on on

[00:13:13] resumes as a as a predictor they also found one other resumes screener that had predicted that

[00:13:20] if you had the word baseball on your on your resume you got more points you know as a predictive

[00:13:24] success and if you had the words softball on your resume that you would get fewer points because it

[00:13:30] was a predicted that you wouldn't be successful at this company so probably where there's all coming

[00:13:35] from is from the training data right that companies maybe use resumes of the folks that are currently

[00:13:41] in the job or have you know applied and have gotten an interview with the last year so

[00:13:45] and then you know the hand it over to an AI tool put it ingested into the system and then the

[00:13:50] AI tool looks for statistical significance so maybe the word you know there are lots of people

[00:13:55] at the work Canada on their resume that were successful at the company good for them

[00:14:00] but you know that's not you know any human knows that there's actually not a predictor success

[00:14:04] neither is at the first name Thomas and it really shouldn't be about hobbies because as you can

[00:14:09] tell with like baseball softball there's probably gender discrimination that comes into play here

[00:14:14] right like most men in the United States they they like and maybe play baseball and women prefer

[00:14:19] often play softball so that's sort of the problem and if you have a company that is not diverse

[00:14:25] that maybe has you know some gender hiring biases or problems in the past and we they you know

[00:14:33] it's take a company who has like a gender disparities with more men and you know this kind of

[00:14:38] stuff can easily creep in into the system if it's not closely monitored and you know it's kind

[00:14:44] of a little bit interesting that maybe the vendors didn't find that out themselves like somebody

[00:14:48] from the outside of the company and to find these problems and we see a lot of these tools you

[00:14:53] know they're bought by companies because they're overwhelmed by applications you know the beauty

[00:14:59] of job platforms is like I get so many jobs as a job secret I can just apply applying it's

[00:15:04] wonderful but I think on the other side it has led to companies feeling absolutely overwhelmed

[00:15:10] with millions of applications like I think IBM gets five million applications a year

[00:15:15] Google three I mean it's just like staggering the numbers um so they're won by a technological

[00:15:20] solution that is efficient you know saves the money saves them labor finds the most qualified

[00:15:26] candidates and is buys free so it's definitely efficient and I think it saves a lot of companies

[00:15:31] a lot of labor and money but we haven't really seen any evidence of finds the most qualified

[00:15:35] candidates and that is not biased and that's really problematic because now we're using these tools

[00:15:40] that scale. So wow so yes the thing is like one is the opacity that you're talking about

[00:15:47] it's just a box like I can't see what's in it therefore like that's not healthy correlation

[00:15:54] and cause hourly you're not the same. The huge problem right it's like you know

[00:15:58] autonomous yeah it's you know it's probably correlation right like it statistically

[00:16:04] significant but we all know that if your name is Thomas that doesn't qualify for a job right

[00:16:09] it's not causally related to a job and any human would know that but obviously a I told doesn't

[00:16:16] know that it just looks for a statistical significance so you know there probably were some

[00:16:21] Thomas's in the pile that was successful good for them but if you don't you know monitor these

[00:16:27] tools and like sort of really critically skeptically look at them and keep supervising them

[00:16:32] these these problems can easily come in because you know usually if keeping giving a training data

[00:16:39] of new people and then this kind of stuff can keep coming in. Well so I mean it's like one is the

[00:16:47] quantity of data it has access to so company X and company Y may have very different data

[00:16:54] pools to be pulled from and then there's the quality of the data there's the historical as you

[00:17:03] kind of point out earlier the historical biases of their hiring practices to such as they might

[00:17:10] embed it into the machines. We're going back 40 years to like one of the most foundational

[00:17:20] computational principles garbage in garbage out like now you give it bad data and then you analyze it

[00:17:26] and not a great way you're going to get not a great result but the good news with AI is we can do it

[00:17:31] scale. Yeah and here in lies the problem right that like I do think that like you know one hiring

[00:17:37] manager and you know I'm sure you know we have unconscious bias a human's like we're not very good

[00:17:42] at hiring either and in fact we often feel like we are very objective and fair and and and reality

[00:17:48] we like people who are like a lot of the you know people went to the same school and yaddy yaddy yaddy

[00:17:55] so that is a huge problem too but like one human bias hiring manager can maybe discriminate against

[00:18:02] so many people and I'm sorry you know they those people are being discriminated against in one

[00:18:07] year but and AI tool that is used on millions or hundreds of thousands of resumes that that has a

[00:18:14] problem with like downgrading mostly women and upgrading mostly men. I mean you could have gender

[00:18:19] discrimination against hundreds of thousands of people and I think you know companies do

[00:18:25] like to use the technology they're a little bit afraid what might happen if they come forward saying

[00:18:31] that the technology didn't work that they're from point so you know I've talked to it a lot of

[00:18:35] HR managers said like oh yeah we had like we did this for a couple years we used this tool and

[00:18:40] we found out it doesn't work so we quietly you know like oh if the problem is like we are

[00:18:46] non-the-wiser right we don't know that there's a problem the vendors aren't being pressured to

[00:18:50] to change their tools and make them better and so there's next company then bias the tool right so

[00:18:56] I think that's really the problem that companies are so afraid of class action lawsuits because

[00:19:01] you know if they come forward and say like oh yeah we we've been discriminated to people

[00:19:05] it was you know there's a gender bias against most mostly women that I mean this huge problem

[00:19:11] so I think you know I need to talk to these you know sort of whistleblowers or folks who in the

[00:19:17] know to tell us what what is happening and that's just not good enough because these are like you

[00:19:23] know as you all know it's like it's high stakes decision and matters if I get a job right

[00:19:27] I'm nervous before a job and if you because it could literally change my life right I need it

[00:19:32] most of us need to work to make money and put roof over our heads and feet of children

[00:19:38] but also if I have a job that like makes me happy full fills me that is a big deal our

[00:19:43] identities tighter up you know I spend so much time at work you know I hope it's like

[00:19:47] somewhat rewarding for people so it does matter I know we get you know a lot of people get

[00:19:51] rejected all the time so it feels like oh yeah it's another way to get rejected but

[00:19:55] if I get rejected for my dream job because I'm not the most qualified candidate I'm not qualified

[00:20:00] enough I get it you know whatever you say at totally but if I would find out that I was rejected

[00:20:05] because at the word softball on my resume and I was discriminated because I didn't have the

[00:20:10] word baseball on I would be in which has nothing to do with the job I would be upset like that

[00:20:14] is not fair okay so let's talk for a minute about solutions the problem seems pretty profound

[00:20:21] if you were in front of you know sure or some large HR organization and we say guys like

[00:20:28] we can all agree that like these are some of the issues what were the top one two three things

[00:20:33] you recommend to them yeah so I mean I think there's there's a bunch of things that we should

[00:20:38] do I think first is like radical transparency and explainability you know it's a problem when

[00:20:44] companies use deep neural networks and you can immediately forget what that means but it means

[00:20:49] that we use training data to train an AI tool but we don't necessarily know as humans what

[00:20:56] the tool then in first upon so we don't know does it use these and these keywords and not all vendors

[00:21:03] check like there is a complicated way to check so I think that's a real problem if we build tools

[00:21:10] where we don't know how the tool um sort of you know judges applicants that is a real problem

[00:21:19] so I think that that we the first thing like transparency and explainability how was somebody

[00:21:25] judge not only were they judged by you know what's their new actual in the in the mix because

[00:21:31] they do feel that most consumers are sort of um and most drop at applicants I call them like

[00:21:36] force consumers because if you want the job and I send you a link to do a one-way video and

[00:21:40] a video with AI you're gonna say no no you want the job so you don't really have a way to say no

[00:21:46] to this so I mean I'm glad if the company tells you that AI is used but it's actually not super helpful

[00:21:52] but explaining how were you judged by the AI or how was their results in third I think that could

[00:22:00] be really really helpful for researchers and others I also think we need like some sort of

[00:22:05] regulation here that like um puts like guard rails up for like some of these high stakes decision

[00:22:10] making so that companies need to make sure that they don't do any harm and more than what we see

[00:22:15] maybe right now in in New York City which is like kind of one of the only places that has a pretty strong

[00:22:20] law here or strong gishla I'll want it to have a strong law to say that you know companies who use

[00:22:28] AI hiring tools need to be audited once a year so there's a lot of you know it's like it's a very

[00:22:34] loose law so a lot of companies I think get out of it and the audits are not very aggressive

[00:22:38] either so that's another problem and I think in general I think we need to have a much

[00:22:43] bunch larger discussion like how should we hire people like we know resumes are not very

[00:22:47] predictable like should we actually use them in hiring like what are some of the ways that we should

[00:22:53] hire people like for example we see a lot of personality tests um they're also not very predictive

[00:22:58] they may be 5 10% predictive so that means that 95 to 90% of your success on job has nothing to do

[00:23:05] with your personality because you know we are humans like we get to sometimes overcome our you know

[00:23:12] our problems maybe my personalities acts but you know I can work hard against that but you know

[00:23:16] these tools can't really um obviously you know like see that or or understand that in a way like you

[00:23:24] know no one can take that into account so I think we really need to think food like what do we

[00:23:28] need to how can we do this better can we have virtual reality to test people on the job maybe

[00:23:33] that is a good way to test folks maybe have a more holistic approach we have different assessments

[00:23:38] you look at the resume you look at a job interview um so it's not just one screen like either

[00:23:44] as a way now you're out like we need to have a much larger discussion here that I'm I hope

[00:23:51] to push a little bit. No I think it's great and then you know on the foot for salt I did an

[00:23:59] interview with the CHR of 130,000 person company recently they hate resumes right because

[00:24:06] the end are desperately trying to find what is the next diversion of you know candidate

[00:24:14] identification and qualification you know where they're way more interested in things like your

[00:24:20] desire to contribute your ability being agile learner your ability to be resilient I mean things

[00:24:26] it really are going to impact your ability to survive and that's survived but even thrive through

[00:24:33] everything that's going on in the world these days and what your job will be and how quickly things

[00:24:37] change so I love all that stuff but we would we need to like really have a scientific discussion like

[00:24:43] how to actually test with that. That's a very hard thing to to measure. It's not easy to build

[00:24:49] if it's any good as a completely different question that's the problem right you know it's a technology

[00:24:54] is out in front of its ability to promise what it really says it can do. Yeah it does look nice.

[00:25:01] I mean my hope is that maybe we'll see like a large starter for something that can do

[00:25:05] maybe some technical assessment or some skills based assessment on sort of hard skills

[00:25:10] because I do think teamwork and agility is actually really hard to to test and we're we're not there yet

[00:25:16] but maybe you know if you're a software developer and you say you know Python you put it on

[00:25:20] your resume as I managed I have no idea you begin or you master developer like where you at

[00:25:25] right it doesn't tell me anything about your skill level so maybe we can do one assessment

[00:25:31] and then it gets like written into sort of the blockchain with a ledger and you have an appeal

[00:25:36] process as a job seeker right and you can like send this assessment to many companies so you

[00:25:41] don't have to redo all these different assessments all the time like it's such a drain on job

[00:25:46] seekers and I'm sure a job manager is not interested in kind of facilitating that either if there

[00:25:52] was like sort of a larger system you know. Are you familiar with some of the tools we're going

[00:25:59] to move on from a TSS in a second but are you familiar with some tools that's basically the

[00:26:04] TSS reverse like job scan and TL has one and yeah yeah yeah I think they're very helpful for

[00:26:11] some job seekers to get like a general sense you know we never know which company has you know

[00:26:18] how is dairy eye tool calibrated like is a calibrated unfokes in the company or is it like an

[00:26:23] off the shelf solution so you know but it gives you a good reading of like okay if you have this

[00:26:28] job description here's the resume how much overlap do you have right and I think this sort of

[00:26:33] general consensus is like maybe 60 to 85% is a good overlap don't do 100% because then some AI tools

[00:26:40] might think you just copied the job description and like throw you around but I think they give you

[00:26:45] a good sense of like what is actually being ingested I think another good signal is it's like if you

[00:26:49] need to upload your resume to some to some companies website and you see that you know maybe your work

[00:26:55] experience isn't in the right column on on the side and when you upload you know usually tells

[00:27:00] it here is like what we thinking what you put in these different columns and feels that might

[00:27:07] be an indication that your resume isn't as well machine readable as it could be so I think

[00:27:13] jobs get and out these other tools can be really helpful because we even see like really you know

[00:27:19] these are semi structured data resumes so computer some of is getting right to get it in the

[00:27:25] deal and I mean the sum of this very basic some of is very sophisticated where problems lie

[00:27:32] so I think that might be one thing for for jobs he could still think through yeah and we do encourage

[00:27:37] people to do that one is your last point around having an ATS friendly resume so it's not complicated

[00:27:43] it's not a piece of art doesn't matter to you like you know if you're just a bit fearful

[00:27:49] but then you know when we're using tools like job scanners or others we we encourage people

[00:27:55] back on the opacity problem we don't know what drives the score but the diagnostics are helpful

[00:28:00] right so you might say customer success and they say client success it's like all right well what

[00:28:06] she's the same vocabulary I mean like if you want to be understood by somebody speak their language

[00:28:10] so what's at least speak the language you make a really good point go about not being like so

[00:28:16] like yeah you did kind of copy the job description and now I mean you should definitely copy

[00:28:21] the most important keywords and it would also suggest not to use synonyms because not all AI tools are

[00:28:27] trained we see sometimes they understand sometimes they don't and I think you know if you want

[00:28:33] a half like a really cool exciting resume you know maybe maybe your machine readable is not exciting

[00:28:39] but if you have then have an in-person job and if you can always bring you know you know

[00:28:44] it's amazing that you love right with the colors and the two columns and you can give that to a

[00:28:49] human but you know machine readable if that is your first encounter it has to be machine readable

[00:28:54] if it's not you kind of already rejected and you know like some other folks who have surveyed

[00:29:00] companies also found out that some of the ATS tools are calibrated to throw out people have

[00:29:06] a six month and longer gap between jobs right and I think that's a really insidious

[00:29:12] problem because you know who knows why he had a six month job that maybe you move maybe

[00:29:16] you're you're sparse in the military maybe he had a kid yet to take care of for his parents or

[00:29:21] were you're a real good person so I think you know knowing that and sort of like you know and

[00:29:28] I think sometimes you just need to just literally cover this gap with like maybe maybe you did

[00:29:33] do some freelance work you know maybe you took care of family members just to put that in there

[00:29:38] so that a machine will put you on the yes pile if you're qualified right because you could be the

[00:29:43] most qualified person if you have that gap and the ATS is calibrated to throw out it will reject you

[00:29:49] so it's like one and out you don't get to like check with a human being and explain why this may have

[00:29:55] happened and you know so I'm not think about benefits like there are some tools for jobs

[00:30:03] seekers that allow you to do like a video interview oh yeah yeah I think it's a great training

[00:30:09] I mean dropping your resume drop in the job description and then it goes okay you want the

[00:30:13] structure of marketing role at a consumer package that's company great and then it says Bob can you

[00:30:19] walk me through your experience and bring you new products to market or whatever and what when I do

[00:30:24] like about them is will we know from our clients is that there's great value in hearing the words come

[00:30:32] out of your mouth like in being able to practice and again with the exact you know math is on the

[00:30:39] back end of the virtual tool to say hey you really crush that interview or that was a great answer

[00:30:44] but the ability to practice and get some feedback you know we do find helpful um here so

[00:30:51] there are other tools maybe that one or other things that you've seen that where AI is being used for

[00:30:58] the good or that that's from the company side or from the candidate side yeah I mean I think what

[00:31:04] would I so I mean if I think that some job seekers are using uh generator AI right check GPT and

[00:31:10] and others to help them you know maybe help them with writing cover letter polishing the resume

[00:31:15] right especially the English is in the first language I think could be absolutely helpful you know

[00:31:20] it's like sort of a a gram allion steroids you just always check that it doesn't

[00:31:24] invent something right like so you have to be a very closely monitoring chat GPT but I think it actually

[00:31:32] helps also with kind of like interview questions right you can ask chat GPT what are the most commonly

[00:31:37] asked interview question and maybe also you know prompted for some for some answers and sort of you

[00:31:43] know maybe you'll find like great answers to like what are you strengthen weaknesses and you know

[00:31:48] sort of those cliche questions and you can prep for that with these video interview tools so I think

[00:31:53] that could be really really helpful to get over this like maybe pretty unnatural setting to

[00:31:58] like talk into your camera there's no one there but you want to sound excitable and you know

[00:32:02] you know excited about the job and it's kind of like okay um and I think you know another thing

[00:32:08] is that like um you know if there's an AI involved like actually longer answers are better than shorter

[00:32:15] for a AI to actually like calibrate a result um so I think that's helpful to actually sort of like

[00:32:22] maybe tell a little bit of a story about we know that also from behavioral questions like you know

[00:32:26] what is an obstacle you overcome like I think you want to um uh talk about that in depth um for

[00:32:33] for tool to work with you but we know very little how to pull out personality like teamwork out

[00:32:39] of these questions like I'm not 100% sure what the what the science behind that and I think that's

[00:32:44] a place for me to dig a little deeper for you for the uh next journalism series um but yeah

[00:32:52] I think those those tools are really great and I do think that there's also some

[00:32:57] uh things on the company side that is like really exciting so we see some um companies that uh

[00:33:03] vendors that help companies sort of um career lattice or career ladder yeah more like a

[00:33:09] lattice not a ladder and any more because people change jobs and functions so many times but we see how

[00:33:15] you know AI tools you know kind of like LinkedIn could actually help you like your playful

[00:33:19] one job but then the company says the company say I tool tells you like hey you know you have

[00:33:24] these other five um uh key skills would you also be interested in this job in this job so you

[00:33:30] know it could like open up your possibilities and it also inside uh the company for employees

[00:33:37] but you often see it's like sort of this career lattice that it tells you okay you are director now

[00:33:41] people in your position five years have done this in this and here's like the skills that you

[00:33:45] need to learn and I think that can be really helpful um I think the problem is a little bit

[00:33:50] that all of this is based again on resume data which we know is like very limiting yeah you know

[00:33:55] we don't know how well you actually are you know how good you are at this particular skill but

[00:34:00] I think it's just good indication so I think it's actually good use case for for big data