BBCNEWS — BBC News
20260920 104500 UTC
215 transcript segments
Original Broadcaster Captioning (Enhanced)
Data courtesy of The GDELT Project (https://www.gdeltproject.org/), from the Internet Archive TV News Archive.
Machine transcription. Treat it as a searchable index of what was broadcast, not a verbatim quotation record.

[00:00:00] Minister Carney remains as popular popular as he has been and whether whether it pushes the Canadians back
[00:00:02] back to the table with the US., which which they say they are open for
[00:00:06] for negotiations, just ones that that are fair. And of course, the
[00:00:08] the other thing is the trade agreement agreement that's in existence there.
[00:00:11] there. The US Mexico-canada agreement. agreement. It's still in place for
[00:00:14] for the guts of another ten years, years, like the negotiations that
[00:00:17] that have fallen apart are about
[00:00:20] about renewing it beyond that term. term. So notwithstanding the impact
[00:00:23] impact of tariffs, a lot of the other
[00:00:26] other regulations are still in place place and a lot of the pain that's
[00:00:29] that's been experienced so far is
[00:00:33] is actually customer boycotts. I I mean, we saw that with the alcohol
[00:00:36] alcohol as well before there was
[00:00:37] was ever a mention of tariffs. Canadians
[00:00:40] Canadians just stopped buying and
[00:00:43] and the shops stopped stocking Californian Californian wine and Kentucky bourbon
[00:00:46] bourbon and all of those things. things. They sort of voted with their their feet. I think unfortunately
[00:00:50] unfortunately that's all the time time we have for now, even though though this is obviously a very meaty
[00:00:54] meaty topic. Indeed it is. And we we will be talking about it again.
[00:00:58] again. Absolutely sure. And convinced convinced of that? Well, great to
[00:01:01] to talk to you guys as always. Talk
[00:01:04] Talk soon. Bye-bye. And you can catch
[00:01:14] catch up with all of the episodes episodes of The President's Path
[00:01:17] Path on BBC iPlayer. If you're in in the UK. Or on YouTube. If you're
[00:01:20] you're anywhere else in the world. world. And if you want to get in
[00:01:23] in touch, please do. You can email
[00:01:26] email us at Pallab
[00:02:28] Gdansk. Join me, Julie Lyn, for discovering discovering the world's table Poland
[00:02:32] Poland on BBC News. With BBC News
[00:02:38] News I'm Lucy Grey. In a moment UK UK viewers will be taken to the Liberal
[00:02:42] Liberal Democrats conference in Brighton. Brighton. But for the moment we're
[00:02:44] we're going to talk about AI and
[00:02:47] and AI is making itself felt in many many different parts of our lives,
[00:02:50] lives, including looking for a job job or hiring someone for a job.
[00:02:54] job. Right now there's what's being being called an application explosion,
[00:02:58] explosion, which is creating challenges challenges for job seekers and employers.
[00:03:00] employers. Our analysis editor, Ros Ros Atkins has looked into the issue.
[00:03:04] issue. Looking for a job is hard, hard, but is AI making it harder?
[00:03:08] harder? Feels like I'm talking to to a robot. It feels really disheartening.
[00:03:11] disheartening. The survival of the the fittest now really. AI stands
[00:03:14] stands accused of ruining the job job market. Do you buy that? No,
[00:03:18] No, I don't. But has a doom loop loop been created where AI is used
[00:03:22] used to write more and more applications applications and where employers
[00:03:26] employers increasingly use AI to to sift through them? Two out three.
[00:03:29] three. Of our clients are really really struggling with the volume. volume. They can simply go click,
[00:03:33] click, click click. If you do get get through the AI shortlisting,
[00:03:36] shortlisting, guess who might be be doing the interview? Once I had
[00:03:39] had this AI tell me that I lacked
[00:03:41] lacked empathy. It's like you're you're speaking to a robot essentially,
[00:03:45] essentially, and I don't think that's. that's. What you are. Well, you are
[00:03:48] are exactly. Recruitment is being
[00:03:50] being rewired by technology, but
[00:03:52] but is it better either for the applicant applicant or for the employer? To
[00:03:56] To help understand the issues, I I wanted to hear from people in the
[00:04:00] the middle of this. Kate runs an
[00:04:03] an AI consultancy. Matt is a recruitment recruitment executive and Niamh Adi
[00:04:07] Adi and Santina are job hunting and
[00:04:11] and sending a lot of applications. applications. Over 200 now about
[00:04:14] about 50. It's been 100 so far. That's That's just in the last year. This
[00:04:17] This is part of an application explosion
[00:04:20] explosion which is happening for for several reasons. Competition
[00:04:24] Competition is fierce. The number number of vacancies is down. The
[00:04:27] The number of graduates is at a record
[00:04:29] record high and AI is helping to
[00:04:31] to write more and more applications.
[00:04:34] applications. 20 years ago, for graduate graduate vacancies, there were on
[00:04:38] on average 38 applications per job.
[00:04:41] job. Last year it was 140. Let's
[00:04:43] Let's look at this from both sides. sides. The employers and first the
[00:04:47] the job seekers. You're trying to to get a job at the moment you're you're finishing university. How
[00:04:51] How central is AI to doing that? that? It's really central. I use
[00:04:55] use it now for finding the jobs itself. itself. When I'm searching for jobs,
[00:04:59] jobs, I have my automated agent that that goes scrapes through the whole
[00:05:02] whole internet seeing which roles roles apply, opening which ones the
[00:05:05] the deadlines are, and it helps me me narrow the selection down. It's It's maybe putting keywords, see
[00:05:09] see which ones can make sound better. better. Even when I know I want to
[00:05:12] to portray a certain side of myself myself and what I want to say and
[00:05:15] and AI can help me sound better, better, maybe edit my CV. I would
[00:05:19] would make sure that it has the same same words that there are in the
[00:05:23] the job description. Adi Niamh and
[00:05:26] and Santina aren't alone. One survey
[00:05:28] survey found 73% of students and and graduates use AI at some stage
[00:05:32] stage during the application process. process. At Western is a senior executive
[00:05:36] executive at recruitment firm Robert
[00:05:38] Robert Half. It places over 100,000
[00:05:41] 100,000 candidates a year globally. globally. It's hard. To find the
[00:05:44] the true skills of the individual. individual. I think with AI you can
[00:05:48] can put in a job spec, it can generate generate a CV. I'm trying to imagine
[00:05:51] imagine the scene in your office. office. Are you sometimes sitting sitting there just going, goodness,
[00:05:54] goodness, more and more and more more applications coming in all looking
[00:05:57] looking quite the same. They are
[00:05:59] are very samey and that's really,
[00:06:02] really, really hard to. That's the the art of what we do. The art is
[00:06:05] is just justifying or really going going through and seeing who is the
[00:06:09] the right candidate. A recent Ft Ft article explained why recruiting
[00:06:13] recruiting is going retro in the the age of AI, when every application
[00:06:17] application is polished. Polish tells tells you nothing. One HR executive
[00:06:21] executive says. In other words, even even if technology helps, human judgement
[00:06:25] judgement is still an essential part part of the process. At this point,
[00:06:29] point, it's important to be clear
[00:06:30] clear on something that the use of
[00:06:33] of technology to shortlist candidates candidates isn't necessarily seen
[00:06:36] seen as a bad thing. The Information Information Commissioner says these
[00:06:39] these tools can enable employers employers to process high volumes
[00:06:42] volumes of applications consistently consistently and quickly. Or listen
[00:06:46] listen to Kate Young. Her business business works with companies which
[00:06:49] which use AI recruitment tools. What
[00:06:52] What we do know is that humans making making decisions about who to hire
[00:06:56] hire is a deeply flawed process, process, and it always has been. been. I'm sure we can all remember.
[00:06:59] remember. Think of a time we were were rejected for a job. We didn't
[00:07:02] didn't have a firm handshake. We We didn't make eye contact or some
[00:07:06] some other spurious question. The The interviewer asked us that actually
[00:07:08] actually nothing to do with our capabilities.
[00:07:10] capabilities. Kate offers a useful useful reminder that humans haven't
[00:07:14] haven't always got it right. Generally, Generally, if the right guardrails
[00:07:18] guardrails are present, evaluate evaluate what it's been told to and
[00:07:21] and assuming those things are job job relevant criteria, there is the the possibility of better outcome
[00:07:25] outcome because all that error and and noise and bias can potentially potentially be screened out, but
[00:07:28] but only if it's built well. There's There's another concern. You might
[00:07:32] might think that applying to ten ten different companies gives you
[00:07:35] you ten chances, not necessarily. necessarily. If different employers
[00:07:39] employers are using the same or very
[00:07:41] very similar algorithms, the same same candidates could be favoured
[00:07:45] favoured again and again. Researchers Researchers have a term for this
[00:07:49] this risk algorithmic monoculture. monoculture. That's not actually actually a new phenomenon. It has
[00:07:52] has not a lot to do with AI. To be
[00:07:54] be honest with you. So if you think
[00:07:57] think back 3040 years, we would typically typically see the same types of people
[00:07:59] people getting the best jobs anyway, anyway, you know, those people were
[00:08:02] were well off. They might be members members of certain majority groups,
[00:08:05] groups, they might have access to to certain social privilege. They They might have gone to the best
[00:08:09] best universities. And those people people would always have had their their pick of jobs. And then if you
[00:08:12] you think about assessment before before AI, employers were still able
[00:08:15] able to configure systems to choose
[00:08:17] choose certain types of people. This
[00:08:19] This is contested territory, both both which biases humans may have
[00:08:23] have and which factors may offer offer reasonable grounds to favour
[00:08:26] favour or reject a candidate. And
[00:08:29] And though it's claimed AI can move
[00:08:31] move beyond human biases given it
[00:08:33] it learns from human data that is
[00:08:36] is far from guaranteed. Whatever Whatever the merits, if you do get
[00:08:39] get through an AI shortlist, next
[00:08:42] next may come an AI interview. Hi. Hi. I'll be your AI interviewer today.
[00:08:46] today. My goal is to learn more about about you and the experiences, skills
[00:08:49] skills and competencies you might might bring to this role. Sometimes
[00:08:53] Sometimes an AI avatar will ask you
[00:08:55] you questions and then respond to to your answers. Other times you're
[00:08:59] you're given a set series of questions
[00:09:01] questions with each answer recorded
[00:09:03] recorded as a timer counts down.
[00:09:05] down. The World Economic Forum says
[00:09:07] says AI led interviews can expand expand opportunities for overlooked
[00:09:10] overlooked candidates, reduce recruiter
[00:09:12] recruiter bias, increase efficiency efficiency and make rejection more
[00:09:16] more constructive. But the experience
[00:09:19] experience of these interviews isn't
[00:09:21] isn't always so positive. It's really really uncanny. It's really inhumane.
[00:09:25] inhumane. And it's like I'm trying trying to speak my heart out, be
[00:09:27] be as passionate as I can. But the
[00:09:31] the persona only sees, oh, are you you able to do this? Are you saying
[00:09:34] saying these key words? How does does it feel when you're talking
[00:09:38] talking to a AI generated person person to try and get a job? It doesn't
[00:09:42] doesn't feel nice. It doesn't feel feel nice. It feels like I'm putting
[00:09:45] putting all this effort in and I
[00:09:48] I don't see like, like with a normal normal human, you'd see. Yeah. Like,
[00:09:51] Like, you know, smiles. Got to the
[00:09:54] the stage where I had an interview
[00:09:56] interview where I was recorded, but but I was essentially talking to
[00:10:00] to an AI and the AI would basically basically ask me questions based
[00:10:04] based on what I was saying. Once
[00:10:06] Once I had this AI tell me that I
[00:10:08] I lacked empathy. So that for me
[00:10:11] me was a bit was a bit crazy. Does Does it matter to you that there's
[00:10:15] there's no evidence of a human being being involved? Yeah, definitely.
[00:10:18] definitely. It feels like I'm never
[00:10:21] never going to be understood by the the other side. It feels like I'm
[00:10:24] I'm talking to a robot. I've got got to be honest. It feels reasonably reasonably bleak that you're not
[00:10:28] not just talking about wanting to to get a job. You're talking about about wanting to reach the point
[00:10:32] point where you can speak to someone. someone. Which is even before getting getting the job, which is even more
[00:10:35] more like, oh, this is such a long long process I have to go through
[00:10:38] through to even get my potential potential seen. Question here is
[00:10:42] is how much of this is the fault fault of new technology? That is
[00:10:45] is a very real and valid feeling, feeling, and I have a lot of sympathy
[00:10:49] sympathy with that perspective. Of Of course we all want to meet a human. human. We all want to feel that connection.
[00:10:53] connection. We want to feel that that our time has been valued and
[00:10:56] and our efforts have been valued valued because applying for a job job is emotional and important. I
[00:11:00] I think it's easy to blame AI and and AI in hiring has its flaws, but
[00:11:04] but I don't think that's the reason reason people aren't necessarily
[00:11:06] necessarily reaching human being.
[00:14:10] learn to live with it, but will not
[00:14:12] not see our lives dominated by it.
[00:14:18] it. In the 24 hour news cycle, patients
[00:14:21] patients can be so frustrating, but but it's a crucial tool. I don't
[00:14:25] don't rush to be the first. I strive
[00:14:27] strive to get it right. I'm Lucy
[00:14:29] Lucy Hockings on BBC News Now.
[00:15:09] Hello. Live from London. This
[00:15:12] is BBC News. In Germany polls are
[00:15:14] are open in two states. Political
[00:15:16] Political pressure is mounting on on Chancellor Friedrich Merz. Two
[00:15:20] Two people are reportedly killed killed after Ukraine carries out
[00:15:23] out large-scale drone attacks on
[00:15:25] on the Moscow region. The brother brother of Princess Diana Earl Spencer
[00:15:29] Spencer tells the BBC that there there are echoes of what happened
[00:15:32] happened to his sister in the way way the media is treating Prince
[00:15:35] Prince Harry and Meghan. Ed Sheeran Sheeran apologises for what he calls
[00:15:39] calls mistakes as he returns to the the stage for the first time since
[00:15:42] since rapper Macklemore was dropped
[00:15:44] dropped from his US tour over pro-Palestinian
[00:15:47] pro-Palestinian remarks. What is
[00:15:48] is happening in Gaza is a catastrophic
[00:15:52] catastrophic and unjustifiable and and disproportionate. My heart has
[00:15:55] has been broken by the scale of devastation
[00:15:57] devastation and loss of civilians