You know when you’re road-tripping with your friends? Wind in your hair, the object of your prospective affection at flusteredly close proximity, youthful abandon running as reckless as the rolling riffs of the song on the radio? And the chorus builds to a crescendo, and you hear the familiar twang of that track you love, and you absolutely BELT out the start of a lyric- …but you came in far too soon, and far too loudly.
And now you’ve just kind of suddenly yelled out into the ether and everyone in the car is looking at you like you’re slightly deranged. And the vibe is weird. You’ve created a core embarrassment memory that you’ll have the delight of unwrapping over and over again every time you can’t sleep.
Now imagine that same mistake, except instead of four friends, it’s in front of the entire world. And instead of just the chance at a holiday romance, you’ve lost out on 100 billion dollars in share value. That’s exactly how some poor press guy at Google is feeling. And it’s all because someone went and jumped the gun in an attempt to get ahead of the ChatGPT-led AI search boom.
What is ChatGPT?
By now, we imagine you will have heard of “ChatGPT.” It’s an AI software, heavily backed by Microsoft, that’s been taking the internet by storm since November 2022. Free for public online use, this “chatbot-style” search engine uses information gathered from hundreds of thousands of internet sources and processes it in real-time to create AI-generated “conversation” with a user.
This can range from simple conversation to complex, tone-directed research information in the format the user requests. In fact, ChatGPT is capable of passing the US Legal Bar exam, so it’s at least as clever as a somewhat competent attorney. To further elaborate (and solidify our nascent feelings of being made horribly redundant by computers), we asked ChatGPT to tell us for itself why it’s such cool software. Here’s how that interaction went:
FGX: “What makes CHATGPT so useful? “
ChatGPT: “ChatGPT is a useful tool for many reasons:
- Natural Language Processing (NLP) capabilities: ChatGPT has been trained on a large corpus of text data and can understand and generate human-like language, making it ideal for tasks such as language translation, question answering, and text generation.
- Versatility: ChatGPT can be used in a variety of applications, such as customer service, content creation, and data analysis. It can also be fine-tuned for specific use cases, making it even more useful.
- Speed and scalability: ChatGPT can process large amounts of data quickly, making it an efficient tool for large-scale projects. Additionally, it can be integrated into different platforms and systems, making it easy to use and scalable for different businesses and organisations.
- Robustness: ChatGPT has been trained on a diverse range of text data, making it capable of handling a wide variety of tasks and languages. This makes it a reliable tool for businesses and organizations that need to process large amounts of data quickly and accurately.
Overall, ChatGPT’s combination of NLP capabilities, versatility, speed, scalability, and robustness make it a useful tool for many businesses and organisations.”
See? Straight from the robot’s mouth. We (scarily) couldn’t say it better ourselves. ChatGPT is the most successful example of this, and it stands to reason that other major companies (like Google, in this case) were looking to emulate this success. And y’know, not really getting there…
So how does Google fit into all of this?
Conversational AI has been around for a while now. Remember the spate of well-meaning AI Twitter bots that inevitably ended up turning into Nazi-propaganda-spewing nightmare fuel? Because it’s Twitter, and what did you expect?
“Tay” went from “humans are super cool” to full nazi in <24 hrs and I’m not at all concerned about the future of AI pic.twitter.com/xuGi1u9S1A
— gerry (@geraldmellor) March 24, 2016
AI has come a long way from us bullying Tay, the Microsoft talking head, into hurried deletion by horrified developers. ChatGPT, backed heavily by Microsoft, is just the most recent, publicly available UI to capture the public’s attention. Google (owned by Alphabet) has been taking exponential strides in AI development for the past 5 years since the creation of the Google AI division in 2017 by then-CEO Sundar Pichai.
The division’s notable projects have been Sycamore (a programmable quantum processor), LaMDA (a family of conversational neural language models) and Magenta (devoted to exploring the role of machine learning as a tool in human creativity, a deep learning research team has developed and launched a series of open-source projects to allow creatives to expand and express their creative process.)
Unfortunately for Google, however, they didn’t jump on the trend in time to be at the forefront of integrating AI conversational functionality with their search engine. ChatGPT cornered the market early and to thunderous applause from the press and high-school students struggling with creative writing essays the world over.
This effectively left Google without the market share in a niche of a market they pretty much created, and one would imagine this made at least one man in a very expensive suit less than pleased. Therefore, Google decided to launch their own solution to ChatGPT, Bard. And Bard missed the mark publicly, to say the least.
So… what actually happened here?
You know when you’re in a chaotic hurry and your jacket gets stuck on a door handle? The door handle in Google’s case is one of the fundamental issues with AI-based information generation.
See, AI gathers its conclusions from the information at its disposal and then repackages that data using language engines that is then sent to you in a neat little bow.
This means that the final informational product you get is an amalgamation of thousands of other sources, BUT there is no guarantee that the information in these sources is 100% factually correct. If you feed an AI garbage source info, you’re not getting a gourmet end result. And this exact problem was reflected in Google’s demo of “Bard” on the 6th of February 2023.
In an official promotional video launched on Google’s Twitter feed (which is inexplicably still live), a user asks the chatbot “What new discoveries from the James Webb space telescope (JWST) can I tell my nine-year-old about?” In its answer, Bard gives a series of facts, and amongst them, suggests that the JWST was used to take the very first pictures of a planet outside the Earth’s solar system, or exoplanets.
As you might have guessed, this just isn’t true. In what we imagine has been the most expensive “ratio” of 2023 so far, an actual astrophysicist at the US Center for Astrophysics and professional looker-at-planets, Grant Tremblay, confirmed that it was a claim made in error.
How did he know? Because he was on the team that produced these interplanetary images in 2004, 14 years before the JWST was even launched.
Not to be a ~well, actually~ jerk, and I’m sure Bard will be impressive, but for the record: JWST did not take “the very first image of a planet outside our solar system”.
the first image was instead done by Chauvin et al. (2004) with the VLT/NACO using adaptive optics. https://t.co/bSBb5TOeUW pic.twitter.com/KnrZ1SSz7h
— Grant Tremblay (@astrogrant) February 7, 2023
Multiple scientists followed suit, calling out the obviously false news produced by this multi-million dollar project, effectively making Bard look like a bit of a twit. By the following day, investors had wiped more than $100bn off the value of the search engine’s parent company, Alphabet. When approached for comment, a rather sheepish Google employee offered that “this highlights the importance of a rigorous testing process, something that we’re kicking off this week with our trusted tester program.” The spokesperson added that “we’ll combine external feedback with our own internal testing to make sure Bard’s responses meet a high bar for quality, safety and groundedness in real-world information.” Unfortunately, though, the damage to the software’s reputational reliability has been done, and its chances of usurping the AI chat search crown from His Majesty ChatGPT are looking slimmer by the minute.
Senior software analyst at DA Davidson, Gil Luria, explained: “While Google has been a leader in AI innovation over the last several years, they seemed to have fallen asleep on implementing this technology into their search product.” She elaborated that “Google has been scrambling over the last few weeks to catch up on search and that caused the announcement to be rushed and the embarrassing mess-up of posting a wrong answer during their demo.” Talk about jumping the gun, right?
What can we learn from this?
Firstly, that the first photographic imaging of a planet outside our solar system was created by Chauvin et al. (2004) with the VLT/NACO using adaptive optics. Yay space!
Secondly, that rushing big releases and campaigns is bound to end in – often costly – disaster. There are a few crucial steps that Google had missed here. One of them was consistent competitor research in their AI division, which would have spurred them on to get developing Bard (and its promo videos) much further in advance.
Furthermore, by not fact-checking their marketing material to protect against an error that they already know might be present in their data, they called into question the reliability of their product and their brand trust. Placing them even further behind their competitors at Microsoft in the race for brand relevance.
Lastly, they didn’t exactly come out swinging with the damage control. The tweet is still up, and those scientists who came forward to dispute Bard’s claim have not been directly replied to or issued an apology for their galactic thunder being stolen. The Google community management team didn’t jump on the opportunity to lean into the meme and make light of their shot in the dark, which always helps to diffuse tense mess-ups.
Now, we’re not saying your business is Google (unless we REALLY nailed the SEO on this article, in which case, love your work, guys.) We’re also not saying our approach to marketing will completely eradicate any and all potential for avoidable obstacles. We are however, here to offer you competitor research and analysis insights, carefully-considered campaign creation and initialisation based on factual consumer data, and diligent community management (with a sprinkling of a sense of humour, to taste.)
Digital marketing can be a minefield, but when an industry leader like FGX is leading the charge, you’re much more informed on the safest places to put your feet. Stop costly mistakes and being left in the dust before you even knew the race has started. Set up a consult with our team today, and let’s see how we can make you look effortlessly cool.
We even asked Chat GPT to write a Shakespearean sonnet to convince you:
“With pen in hand and heart full of rhyme,
I write to thee, of a firm most bright,
In digital realm, its fame doth climb,
With power to bring forth thy business’s light.
FGX, the name, a firm of great renown,
With expert skill, in marketing their trade,
They’ll make thy brand, in beauty, brightly known,
And bring to thee, success, as a lavish parade.
Their methods, tried and true, a flawless art,
A tailored touch, that doth bring great gain,
Thy business’s name, they’ll set apart,
And in the minds of all, a lasting stain.
So trust in FGX, do not delay,
For they shall guide thee on success’s way.”