Chat GPT is like a cross between an encyclopaedia, Google and a witty writing tutor

 Mike O'Donnell is a professional director, writer and strategy adviser, and a regular 

opinion contributor.

OPINION: My old mate university mate Chris Thompson has a colourful turn of phrase. Along with a handful of digital troublemakers Thompson drove the exponential growth of web portal Xtra in the late ‘90s at Telecom.

Along with maverick leader Rod “Snoddy” Snodgrass, he proved skilled at keeping the corporate mothership at arms’ length often with a cutting quip or two.

Last week Thompson told me that he had nicknamed a workmate “ChatGPT” because the bloke states the obvious quickly but never arrives at a recommendation for action.

READ MORE:* Google owner Alphabet faces share price wipeout after Bard AI chatbot gives wrong answer* AI chatbot race between Microsoft, Google is loaded with risk* Can ChatGPT review restaurants? One food critic tests it out

The quip sums up how many would see the chatbot that’s taken the world by storm since its launch a scant 10 weeks ago.

Chat Generative Pre Trained Transformer (ChatGPT) was built by United States artificial intelligence research company OpenAI and has quickly gathered global acclaim for its remarkable responsiveness and insightful answers to questions.

Think of it as a cross between an encyclopaedia, the Google search engine and a writing tutor with the personality of Hugh Lawrie, and you start to get a feel for ChatGPT. You can ask it to deliver a haiku on the subject of burning dumpsters floating down a river, and it will do so in about 5 seconds flat.

Kevin Stent/Stuff

Mike O'Donnell says the sheer speed and charm of ChatGPT leaves him breathless.

Ask it to deliver a 500-word essay on the relationship between Mexican artists Frida Karlo and Diego Rivera and about eight seconds later one will pop out. And it will be remarkably good. Grammatically it will be perfect. A fact or two might be slightly shaky.

But it’s getting better.

Before Christmas I asked it how to tune the engine on a 1970 Volkwagen Kombi, and it gave me a generic but articulate version of how to get fuel air mixture right and advance timing.

I tried the same thing again yesterday and the advice was customised and pretty much perfect; so now my 1778​ twin port donk is running smoother than ever.

But now the achievement bar is being pushed higher and higher for this v1.0 enfant terrible of the deep learning world.

In January it managed to score a passing score on the American medical licensing exam for general practitioners. And last week it passed the post graduate law exams for the University of Minnesota’s law school.

And it keeps on going and getting smarter. And some would say funnier.

It’s so good it scared the bejeepers out of Microsoft, so they threw US$10 billion (NZ$16b) into parent company Open-AI which is now valued at US$30b. Not bad for a 380-person company that’s just 7-years-old.

Mind you, initial investors at OpenAI include Peter Thiel, Vicki Cheung and Elon Musk – not exactly a bunch of clods.

It’s put the wind up Alphabet enough, that they declared a “code red” last year given it could threaten the global dominance of the Google search engine.

No surprise then that just a few days ago Alphabet head honcho Sundar Pichai announced the launch of Bard, an experimental conversational AI service that looks remarkably like ChatGPT. Rumour is that the team at the Googleplex threw an extra 1200 people onto the project late last year and worked right through Christmas.

Alphabet is already promising that Bard will be more widely available to the public in the coming weeks.

For the last few years, everyone has been very aware of the rise and rise of AI as a game changer for virtually every business. Eighteen months ago I wrote about data scientists being the weapon of choice as the world moved from predictive analytics, to machine learning to deep learning.

But the sheer speed and charm of this latest manifestation leaves me breathless.

The black hats have been keen to point out it's not all beer, skittles and cute anecdotes. Given it’s essentially an open-source chatbot, the code is open to illicit modification and a hacked version could fuel very targeted cyber-attacks.

As easily as it can give you killer recipe for scones, it can give you a killer recipe for phish mailing, pharming and API attacks. Plus given its thirst for data sifting, a modified version of the code could be used to steal data. A lot of data.

All of which perhaps explains why this week ChatGPT co-creator and chief technology officer at OpenAI Mira Murati has come out in Time magazine saying that she reckons there is a narrow window where people have some say in how ChatGPT and its ilk will shape society. And that we need to make the most of that window.

She even got dangerously close to saying that governments should be involved in ensuring its adoption is controlled and responsible.

Most technology thinkers tend to agree with John Perry Barlow’s original Declaration of the Independence of Cyberspace. Namely “governments … leave us alone. You are not welcome here. You have no sovereignty”.

So for Murati to be advocating for the opposite is a first as far as I can remember. And a chilling reason to ensure we heed her advice while we still can.

I just hope governments are all listening. And specifically that, like Thompson’s colleague, they don’t just state the obvious without taking a recommended action in this narrow window

Google employees criticise CEO Sundar Pichai, term announcement of Chat GPT rival Bard as ‘rushed, botched’

Seems some Google employees are upset with the way the company announced its ChatGPT rival Bard. According to a report in CNBC, some Google employees are criticizing leadership, most notably CEO Sundar Pichai, for the way the company handled the recent announcement of its ChatGPT competitor called Bard. These employees took to the popular internal forum Memegen to express their thoughts on the Bard announcement. Some calling it as “rushed,” “botched” and “un-Googley,”. Messages float on internal message forum“Dear Sundar, the Bard launch and the layoffs were rushed, botched, and myopic,” reportedly reads one meme that included a serious picture of Pichai. “Please return to taking a long-term outlook.” “Sundar, and leadership, deserve a Perf NI,” read another post, refering to the company’s employee performance review system. “They are being comically short sighted and un-Googlely in their pursuit of ‘sharpening focus,'goes another post. Some messages also referred to the 12000 job cuts annouced by the company in January this year. The meme featured a photo of actor Nicolas Cage smiling and said “Firing 12k people rises the stock by 3%, one rushed AI presentation drops it by 8%.” Another showed an image of a dumpster fire with the Google “G” logo on it. The text said, “How everything’s felt since last year.”Earlier this week, Google announced “Bard” to compete with ChatGPT in the large language models (LLMs) space. CEO Pichai made the announcement via blog post. Bard is an “experimental conversational AI service powered by LaMDA” — Language Model for Dialogue Applications that notably “draws on information from the web to provide fresh, high-quality responses” instead of having its training data capped to a certain year.Google has rolled out Bard for “trusted external testers”. Wider, public availability will be in the “coming weeks.”