HaoTech
KEY INSIGHTS
The release of Google’s Gemini, touted as the most formidable AI to date, has stirred a mixed response – from resounding acclaim among some users to skepticism and queries among tech savants, signaling a potential shift in the AI landscape.
It’s premature to gauge the eventual impact of Gemini, Google’s latest AI offering, on the world – Is it the new ChatGPT conqueror or just another contender? However, initial responses are shaping the narrative.
As anticipated in early November, the distinguishing factor of Gemini lies in its multimodality. Google certainly lived up to the expectation by unveiling a next-gen AI capable of comprehending and manipulating a diverse range of content, including text, images, audio, video, sound, and speech.
Yet, the AI – hailed as the world’s most potent – didn’t receive unanimous, unqualified praise.
While some users are all praise for Gemini, proclaiming its superiority over Chat GPT-4, tech pundits have raised numerous pertinent questions. It’s clear, the jury is still out on Gemini.
Google Makes a Resounding Comeback,” Users Declare
Users can only get a taste of Gemini through Bard, Google’s renowned free online chatbot. Google asserts that the Gemini Pro version integrated into Bard endows the chatbot with enhanced code generation, advanced image and video comprehension, and cross-modal reasoning capabilities (meaning it can amalgamate information from text, images, and videos to produce more precise and pertinent responses).
Google’s Gemini: The Three Facets
To some users’ disappointment, Google also disclosed that the more superior version, Gemini Ultra, is slated for release sometime in 2024. The company also fell short of delivering Gemini Nano, the AI designed for mobile devices, which is projected to be launched before the year concludes. Despite this phased release approach, numerous users took to social media to voice their admiration for Gemini.
Business Insider reported a flurry of “Google is so back” tweets following Gemini’s launch, countering the prevailing notion that Google had lagged behind competitors like OpenAI. Some users even boldly claimed that Gemini is “so much faster” than GPT, with others sharing test results on social media to demonstrate Gemini’s supremacy.
Another feature that has caught users’ fancy was showcased in a Google YouTube video that quickly went viral, amassing 1.3 million views in less than 24 hours. The video, showcasing Gemini’s video-speech AI capabilities, was inundated with comments like “blew my mind” and “impressive.”
Experts Cast a Discerning Eye over Google’s Gemini
While users showered praises on Gemini and various global media outlets echoed Google’s press release data about Gemini surpassing GPT-4, a number of tech experts treaded the Gemini hype path with caution.
In a report entitled, “Google’s Gemini isn’t the generative AI model we expected,” TechCrunch stated that they couldn’t independently verify if Gemini Pro enhanced Bard. The outlet added that Google neither allowed the press to test the models prior to public release nor provided live demos during the Gemini press briefing to corroborate these claims.
TechCrunch reported that Gemini “isn’t much to write home about,” and that the advanced model, Gemini Ultra, was “somewhat more impressive” than the Pro version. Although no one disputes that Gemini’s multimodal abilities surpass Chat GPT with vision — which only works with images and text — the outlet did raise some serious questions that Google would eventually need to address. Particularly, TechCrunch noted that Google consistently skirted questions about Gemini’s training data.
“Google repeatedly refused to answer questions from reporters about how it collected Gemini’s training data, where the training data came from, and whether any of it was licensed from a third party.”
MIT Technology Review also had some strong words for Gemini. The outlet questioned if the wait was worth it and suggested that despite appearing amazing, “Gemini could signal the peak of the AI hype”.
Like several other outlets that delved into the performance data presented by Google as evidence of possessing the most potent AI model, MIT Technology Review concluded that while Gemini does outdo GPT-4 in almost all aspects, the margin is slight.
“It’s a big step for Google, but not necessarily a giant leap for the field as a whole. Google DeepMind claims that Gemini outmatches GPT-4 on 30 out of 32 standard measures of performance. And yet the margins between them are thin.”
AI: More Than Just a Buzzword
In this report, we won’t delve into AI model performance comparisons — you can find our comparison with ChatGPT elsewhere. It’s evident that Google’s Gemini is just beginning its journey, and in our view, this new AI is both amazing and groundbreaking. Its advanced capabilities signify an invaluable technological milestone.
Yet, a deeper trend seems to be at work. With AI being hailed as the
coveted “Holy Grail” of the tech world, are we to anticipate a showcase of new AI products by major tech giants at every event?
Both tech enthusiasts and laypeople are aware that we’re presented with a ceaseless stream of new smartphones, tablets, VR/AR glasses, computers, gadgets, and devices each year.
These are all pitched to us as the next big thing, even though, more often than not, they constitute incremental advancements rather than disruptive breakthroughs.
Systems are becoming more human-like than ever; they can converse, listen, comprehend, and interpret images and videos and are projected to merge with the real world via robotics. The AI era is upon us. And Gemini is blazing its own trail on this challenging journey.
However, the incessant proclamations of “bigger, better, faster, stronger” AI models are beginning to echo like vacuous marketing slogans.
This overshadows the inspiriting efforts invested by AI developers and the remarkable potential that lies beneath the surface of this transformative technology.
When innovation is synonymous with business, mixed feelings about new technologies are a given.
