Broadstrokes: Eliminating human error
ONE OF THE MOST interesting stories during my recent holiday abroad was that US information service Thomson Financial has devised a way of replacing live reporters with computers to write results stories. It claims it can generate an earnings story within 0.3 seconds of a company publishing results.
According to an article in the Financial Times, by retrieving previous results from its data base, the computer can even say whether the results are better or worse than expected.
A senior Thomson executive was quoted as saying: "This is not about cost but about delivering information to our customers at a speed at which they can then make an almost immediate trading division. This can free up reporters so they have more time to think."
He added that computer-generated stories make no mistakes, but are very standardised. "We might try to write a few more adjectives into the program."
Now the implications of this are tremendous, starting with the simple confirmation of my long-standing belief that most adjectives are redundant. More importantly, elimination of human error is a way of overcoming what is widely perceived as a steady deterioration in the quality of journalism.
A few eccentrics resist the temptation
Only this week a daily business publication, reporting the package of a departed chief executive, confused Johnnic and Johncom, dashing the hopes of those of us still waiting eagerly to learn what Connie Malusi took out when he quit the latter company.
This deterioration is not surprising. Societies get the quality of journalism they're prepared to pay for, and the relative status and financial rewards of a career in financial journalism have deteriorated progressively during my decades in the trade, while those in financial services have correspondingly improved.
Why stay in financial journalism when you can lift your income by 50% or more by going into corporate affairs, public relations or investment research?
A few eccentrics resist the temptation, but far too many talented people spend two or three years in journalism, which remains an unrivalled way of making contacts and broadening one's experience, and then go off where the money is.
With some honourable exceptions - including, naturally, Media 24 - publishers no longer regard journalists as part of their intellectual capital, but as cost factors.
And within publishing, the gap between salaries of journalists and managers has also ballooned. As it's the managers who set the salaries, this shows their view of the value of journalists.
Which is a roundabout way of saying that we can take with a pinch of salt the Thomson executive's assertion that computer generation of stories is not a cost-saving initiative.
Main beneficiaries
As for the suggestion that it will give reporters more time to think ... well, editors may appreciate independent thinking, but their influence vis-à-vis the bean counters has also diminished, who're concerned with margins and productivity rather than the quality of editorial copy.
But it's not only publishers who could benefit from computer generation of stories. The announcement of the Thomson initiative coincided almost to the day with a report that bonuses in the City of London are likely to raise by 8% this year, to a record £21bn.
That figure astounded me so much that I checked with several reports, but it's accurate. Yes, folks, the equivalent of about R300bn will be distributed among a few thousand people. The range is wide, but most can expect anything between double and 10 times their basic salaries.
Corporate finance professionals will be the main beneficiaries, with the boom in merger & acquisitions business; but investment analysts will also be on the gravy train.
Now we know that buy recommendations make up a good 90% of all stockbrokers' published research, whatever the state of the market. Who needs a highflying analyst to write these?
What more do you need?
It shouldn't be too difficult to enhance the Thomson program to generate stockbrokers' research notes, too, saving substantial salaries, bonuses, and all the other costs of employing highly-paid professionals.
Savings here would far outstrip those that could be made in journalism.
You could guarantee that the figures would be correct, and there are already programs that can run earnings projections and assess comparative value.
What more do you need? Completely automatic generation of brokers' notes is just around the corner.
Those human analysts not made redundant could indeed spend more time thinking, looking for special situations and developing contrarian views, and because they would suddenly be in oversupply, would have to accept big salary cuts, generating yet more savings for their employers.
Of course, those pesky adjectives might still be a bit of a problem. But given what's at stake, I'm sure a solution could be found.
News source: www.news24.co.za
Posted by: www.SouthAfrica-CarHire.com
We could automate vehicles too, but wouldnt that just be boring?
Click Click... Vroom, Vroom
