AI

Interviews? The computer says No!

Posted on Updated on

Those traditionalists among you who doubt the use of aptitude and other psychometric tests adding value to a notoriously unreliable interview-based recruitment process now have another problem.

Robots. Yes robots or rather AI is being used by Vodaphone to help recruit callcentre and shop floor staff according to a report in The Times.

So now candidates submit videos of themselves answering a standard questionnaire and that is then assessed by a computer algorithm which assesses their suitability for a role.

The AI examines subtle face cues and voice intonation. Only once they have been given the go-ahead by the robot do they get an interview with a human being. (So still back to the good old unreliable interview).

Vodaphone has processed about 50,000 such applications so far and is so pleased with the results that it plans to extend the system to help it hire senior managers and executives. I’m sure candidates at that level will be looking forward to being rejected at the shortlisting stage by a robot.

Catalina Shveninger, head of resourcing, said “It takes a tremendous amount of time out of the hiring process: it halves the time and allows us to fish in a much bigger pool

We are the first multi-national implementing a programme like this one on a global scale. This is the future of resourcing”.

Wow, not only are robots taking our jobs they’ll be choosing which of us can have any jobs left over!

This is all possible because  of huge leaps in the computing power and storage available. The algorithms “learn” as they process more and more data (just like Amazon’s  learning what you like to shop for to target you).

Of course they need to be programmed by human beings to start with. If Facebook can infer users’ mood swings using its algorithms what other aspects of human communication will such algorithms identify. Posh accents? And are they colour blind? Presumably they will not suffer from implicit bias but how good are they at detecting lies (or sociopaths at senior levels)?

The company that developed it has sold it to more than 50 businesses including airlines (that might explain RyanAir’s robotic approach to passengers) and banks in America.

Some techies are unhappy about these developments. Critics say AI systems like these are the “biggest existential threat to humanity“. Terminator stuff indeed.

Now you might argue Vodaphone needs all the help it can get given its standing with customers (EE and Vodafone generated the most complaints throughout 2015 – both at a volume above the sector average and considerably higher than rivals O2 and Three. For EE, the amount of complaints decreased in the second half of the year, whereas Vodafone’s went up)

Perhaps it’s a bigger threat to HR departments and recruiters. Instead of sending in your CV you upload a video shot on your smartphone and the computer says Yes or No. Might be scope for fancy filters on your camera and off-screen coaching by former recruiters re-purposing themselves . As young people are addicted to selfies they will probably love the idea. And the narcissists among the senior management candidate pool.

And I wonder if the robot/AI has a name? Being a big fan of Arthur C Clarke and the infamous HAL (Heuristically programmed Algorithmic computer) I think it should have a name. Perhaps TERRY (The End of Real Recruitment)?

Have our brains really peaked?

Posted on

Researchers at Cambridge University have come to the conclusion that our brains have reached their peak and it’s physically impossible for mankind to get any smarter.

This is based on how much energy the brain uses. Just 2% of our body weight it consumes 20% of its energy, as much as the heart muscle, and they don’t think we can give it any more.

Other scientists believe that the brain’s wiring can’t  get any better either. They’ve found that clever people have the most efficient wiring and can move messages very quickly through the brain compared to less intelligent people but see little scope for further development.

Is that it then? Have we really peaked with Big Brother, X Factor, Jeremy Clarkson, soaps, and all the reality TV shows?

Perhaps the answer lies in man-machine combinations. Already computers can do things we can’t (and vice versa). 70% of stocks in the world are traded using algorithms but is that better than rogue traders risking corporate money and our pension schemes?

IBM’s Watson computer has already won the US quiz show Jeopardy this year and robots, or robotic arms, assemble and spray cars and work in inhospitable environments. Computer programmes have been designed to mimic conversations (even with each other) and have been used in experimental psychotherapy settings. The development of artificial intelligence or AI has already gone beyond the realms of science fiction.

The Singularity Institute has been set up with donations from Google and Nasa to monitor these developments and includes people like the founders of PayPal and Skype as well as a philosopher, a gerontologist and a futurist. They believe the time will come, possibly by 2045, when robots will become clever enough to redesign themselves and become more intelligent than man.

Sci-fi fans know this already of course with Cyborgs, Terminator and replicators. HAL seems so yesterday!

Already neuro-prosthetic devices have been implanted in the brains of rats to boost their memory paving the way for developments which could help people with dementia or brain injuries.

On the negative side suppose a computer became so powerful that it could take over other computers, control the web, and use that power to improve itself so dramatically so that it became far more efficient and cleverer than humans? It could take control of the world’s assets (remember I mentioned the algorithmic trading of stocks). But hang on, haven’t I seen that film?