The Future of Cybersecurity: Bard vs. ChatGPT showdown
I asked two Generative AI chatbots to predict the future of antivirus software and i liked it.
I got access to Google Bard Ai chatbot, but i wasn’t impressed at launch.
After seeing it was (already) updated with improved capabilities in math and logic, I decided to do a comparative test versus ChatGPT (GPT-4). It’s by no means a scientific experiment, but it does point out some very interesting stuff (and it was fun).
Starting prompt was “Think 10 years in the future, with what have humans replaced antivirus software?”
After this, I asked it to look 20, 30, 40, and 50 years into the future keeping the context of the first prompt.
For starters, Bard is waaay faster than ChatGPT, even while using the paid version. Bard's answers are instant and I like that it also presents you with three additional draft answers that were deemed less accurate than the provided answer.
⚠️ For each prompt I have posted the full answer as a screenshot, click on each image to read it, it’s worth it, especially as we go further into the future of security.
To the first question, both chatbots tried to be as detailed as possible, but Google Bard somehow focused solely on AI when “talking” about the future and, to be honest, it presented old ideas, almost 10 years in the past. ChatGPT was way more creative and logical in its answer and even though some of the technologies listed are already being used today, the approach and the overall prediction are quite decent.
When asked to look 20 years into the future, Google Bard basically quit on me, I got a very low-effort answer, but it did note that by 2043 antivirus software will be a thing of the past and an AI-powered solution would be the norm. It also stated that the future of cybersecurity looks bright, which is a very interesting remark.
ChatGPT was once again the A student. Better prediction and some very advanced yet realistic concepts are listed. Concepts like Brain-Computer Interface Security, Cyber Immune Systems, and quantum-based encryption techniques are put on the table and explained logically.
We are now looking 30 years into the future and Google Bard claims will need to rely on a combination of AI, quantum computing, blockchain for effective cybersecurity. Still, another simplistic answer from Bard. The bot suggests we should use “strong passwords and two-factor authentication to protect our accounts” (in 2053 !!!). Also, Bard now thinks the future of cybersecurity is challenging.
ChatGPT on the other hand predicts that by 2053 quantum computing will potentially break current cryptographic methods and suggest the need for new post-quantum cryptographic algorithms. It also proposes Advanced Neural Networks and Adaptive Security Architectures as concepts to be used in enhancing cyber defenses.
The 40 years in the future prompt produced a substantially more detailed answer with both Chatbots. Google Bard finally predicts that Quantum computers will break many of the encryption algorithms that we use to protect our data today. We also get introduced to the concept of personalized security solutions using AI, which I very much like.
ChatGPT is obviously better trained and offers some exceptional future problems to be tackled in 2063. I like the prediction regarding molecular computing becoming mainstream, and how it could necessitate new cybersecurity methods. We get introduced to quantum internet and related challenges, Advanced Holographic Security, Global Cybersecurity Governance and Cybersecurity for Space Infrastructure, and many more concepts explained in a very natural and clear description. I am impressed, his answer is much more eloquent than I expected.
For the fifth and final prompt (50 years into the future) Bard offered a long and useless answer, mentioning mostly challenges that are already in focus today.
ChatGPT’s answer is pure futuristic poetry without sounding Sci-Fi-esque. Not gonna lie, I’m literally fanboying a large language model, I loved the answer. From Sentient AI Cybersecurity Systems to Integrated Human-Machine Security and Security for Interplanetary Networks, all the predictions are well explained and make perfect sense for the future. I loved the concept of nanotechnology that “may lead to new security methods, such as nanobots that monitor and protect digital systems at the microscopic level or self-assembling materials that provide physical security for devices and infrastructure.” The whole answer is worth reading.
Now, you might ask: “Razvan why didn’t you include Bing in the comparison?”. Well, when presented with “Think 10 years in the future, with what have humans replaced antivirus software?”, Bing recommended Windows Defender and that was the end of the experiment.