Governments struggle to regulate AI, Mixed signals continue with regard to optimism or pessimism in tech spending and Artificial Intelligence is making some employees so productive they can do the job of two people – and maybe get paid for both.
 |
 |
 |
These stories and more on Hashtag Trending, for Friday April 28th. I’m your host Jim Love, CIO of IT World Canada and Tech News Day in the US.
Governments are struggling to control what might seem like the unstoppable force of AI development. They know there will be a massive impact. They know that there are risks. But how do you regulate something that we barely understand in terms of its capabilities today and that is evolving at an unimaginable rate.
And competition between companies and even countries means that we might sacrifice safety for superiority. Can we? Google and others cannot let Microsoft dominate without a fight. Even at the government level, can the US cede the field to China in terms of AI superiority?
There is a growing sense of alarm in the public and in government.
Senator Chris Murphy summed up the government’s dilemma when he said last month, “something is coming and we are not ready.”
In the US, Senate Majority Leader Chuck Schumer is trying to develop legislation to regulate artificial intelligence technology according to reporting from Axios. Schumer is reported to have met with Elon Musk and others in trying to draft a workable framework for regulation.
The National Institute for Standards and Technology (NIST) has put out an AI Risk Management Framework earlier this year. The White House Office of Science and Technology Policy has also published the blueprint for an AI Bill of Rights.
According to a report in Axios, the government approach is consolidating around four principles. The US approach centers around transparency and disclosure:
- Identification of who trained the algorithm and who it’s intended audience is,
- The disclosure of its data sources,
- An explanation of how it arrives at its responses, and
- Transparent and strong ethical boundaries
This is pretty much what experts are looking for, although many want the regulation to go much further than simply principles.
And if the US government seems reluctant to regulate technology, the Chinese government is not. It has earlier issued guidelines on transparency of algorithms and models.
According to a report originally in the New York Times, the Chinese government has issues draft regulations that say that companies must ensure that their algorithms are truthful and respect intellectual property and must register their algorithms with the government. But they go further, and require that:
“Companies must heed the Chinese Communist Party’s strict censorship rules, just as websites and apps have to avoid publishing material that besmirches Chinese leaders or rehashes forbidden history. The content of AI systems will need to reflect “socialist core values” and avoid information that undermines “state power” or national unity.
The article reports that Chinese scientists are already figuring out how to comply with this as the struggle to catch up with the US developments like ChatGPT. But it’s not that easy. For instance, any Chinese AI must know a myriad of things that it can’t talk about.
And if you think that Google’s Bard faux pas where the chatbot got an answer wrong about the discovery of a planet was a big deal, a recent Chinese AI got more than 10,000 inquiries in its first hour, but gave answers that were less than favourable about the Chinese economy and described the Russian Ukrainian conflict as a war of aggression.
The entrepreneur shut it down and said he was working to develop a more “patriotic bot.”
The bot was supposed to resume February 13th according to screenshots that read “Service will resume when troubleshooting is complete.”
As of today, it appears that they are still “troubleshooting.”
Sources include: Axios and The Sydney Herald with a reference to the New York Times.
—
We are still getting mixed and often contradictory signals with regard to the state of the tech industry. We have reported on Microsoft and Google having positive results exceeding analyst expectations, largely from their cloud divisions. Google’s cloud showed a profit for the first time ever.
On the other hand, Dropbox, the cloud storage company has announced that it will reduce its global workforce by 16 per cent to cut costs due to “slowing cloud growth.” Dropbox’s CEO Drew Houston said the company’s core cloud business growth was slowing as the economic downturn was putting pressure on its customers making some investments no longer sustainable.
Dropbox also noted that it will be hiring new resources to build out its AI offerings in an attempt to keep up with the growing competition in that area.
Wipro, the Indian outsourcing giant reported that its growth in the the current quarter could be as low as .06 per cent, “as clients cut down on discretionary spending.” Another of the global Indian outsourcers Infosys reported slower growth in the range more of 4 to 8 per cent as customers in the US and Europe “postpone or cancel deals due to the weakening global economy.
On the other hand, Meta, parent of Facebook, after three quarters of declining earnings saw year over year revenue growth. Meta’s projections were actually that revenue would decline in the last quarter, but it actually increased by 3 per cent, despite continued losses in Reality Labs where its metaverse development continues.
Meta’s revenue growth is largely from a rebound in advertising spending, which is generally as sign of optimism.
Maybe Charles Dickens got it right when he said, “it was the best of times, it was the worst of times.”
Sources include: Reuters (Wipro, Infosys) (Dropbox) and Axios (Meta)
__
Every year there is a new set of studies on passwords, usually talking about how simple and easy to guess they are. Passwords like 123456 and even password hit the number one lists.
A new study from Bit Warden confirms that 45 per cent of people surveyed had easy to guess or simple passwords.
Other bad practices in the report include sharing passwords by email, often unencrypted email or in unencrypted documents. Close to 40 per cent admitted to both of these.
But the shocking number is that 90 percent of decision makers reuse passwords. If there is one thing that makes your password easier to crack that a ridiculously simple one like 123456, it’s reusing a password.
Chances are that almost every password you’ve used in the past has found its way into one of the huge repositories of passwords that are used in so called dictionary attacks where hackers cycle through a list of passwords.
But when you reuse a password that has been compromised, it makes it even easier. Phishing you on Facebook might also get your banking or corporate passwords. That’s why you are always told, presumably by some of these decision makers – never to reuse passwords.
Another finding of the survey – 60 per cent of those surveyed say they have been victim of a cyber-attack.
I wonder if those decision makers surveyed would see any correlation between these things?
Sources include: Bit Warden
And the use of AI is making some employees more productive – so productive that they can take another job.
Vice’s Motherboard interviewed workers who were adopting AI and found someone, they used the pseudonym Ben, who had become so productive using the new release of ChatGPT 4 in his current job that he was able to apply for and get a second job and do both at once.
He occasionally has to tweak the content, but as he told Vice, ChatGPT does, like, 80 per cent of my job if I’m being honest.”
So he applied for a second job and got it – using, of course, ChatGPT to write his cover letter. He said. “That’s the only reason I got my job this year.”
How does Ben manage his two bosses? You don’t need AI to predict this one – he’s using ChatGPT. And he’s trained ChatGPT to write in all lower case, to make his messages to his new bosses seem more authentic.
Once you can fake sincerity, you’ve got it made.
Sources: Business Insider and Vice
That’s the top tech news for today. Hashtag Trending goes to air with a daily newscast five days a week, as well as a special weekend interview with an expert on topics relevant to today’s tech news.
Follow us on Google, Apple, Spotify or wherever you get your podcasts. And you can even get us on your Alexa or Google smart speaker.
We love your comments. You can me on LinkedIn, Twitter, or on Mastodon at @therealjimlove on our Mastodon site technews.social. Or if that’s too much, just leave a comment under the text version at itworldcanada.com/podcasts
I’m your host, Jim Love. Have a Fabulous Friday!