Since the launch of ChatGPT in November 2022, generative AI has become a public debate around the world. According to GlobalData, there were over 1 million social media posts about artificial intelligence (AI) across Twitter and Reddit last year.
The Future of Life Institute’s open letter calling for a moratorium on AI development has led several major news outlets to run special features on the existential threat AI could pose to humanity. ChatGPT’s instant internet virality is probably the biggest advantage so far for the tech industry.
Generative AI is being deployed to provide customer relationship management solutions, software development, and even storytelling. But some skeptics can’t ignore the timing of the generative AI hype for the tech industry.
Dario Perkins, managing director of TS Lombard, explained in a recent webinar that big tech is “the part of the stock market that has suffered the biggest declines as central banks have begun to aggressively raise interest rates.” .
Perkins also attributed the “depletion” of technology investment to the closure of several banks with close ties to the tech sector, including Silicon Valley Bank, and called the public launch of ChatGPT “very smart marketing.” there is
As Metaverse winter approaches and funding for quantum computing slows due to a scarcity of practical applications, generative AI instead begins to experience a funding frenzy, while funding for AI overall declines. Did. On June 13, French company Mistral AI broke the European record for seed round funding, raising a total of $260 million in four weeks.
AI is also driving hiring in tech companies, and Perkins said these companies are creating “pretty big jobs compared to other sectors that are still looking to add jobs. He points out that he is experiencing a loss. According to a GlobalData study, employment in the tech sector’s generative AI field increased by nearly 600% from March 2023 to June 2023.
In addition to the influx of news and social media coverage and increased hiring, a recent research report by MIT found that using ChatGPT increased efficiency by nearly 40%, significantly improving the quality of work for low-performing students. I am proposing that it is possible.
GlobalData’s Technology Sentiment Survey shows that AI will be recognized as the most disruptive technology in the final quarter of 2022. Nevertheless, many companies including Apple, JP Morgan, Deutsche Bank and Verizon have banned the use of generative AI in 2023.
Will a lack of understanding of generative AI hinder its progress?
As search engines emerge to optimize internet use, people wonder if simply “Googling” a question slows down our brains and reduces our ability to retain information. No wonder. This same question is now being asked of AI.
Responding to these concerns, Josep Boli, research director at Global Data, said that current generative AI could be a “very useful and powerful tool” but that “probably soon it will become a fact that it is not fully monitored.” It’s not going to be based advice.”
Another challenge of generative AI is the illusion of large language models (LLM). Bori remains pessimistic that the issue will be fixed soon, citing LLM’s inability to fact-check answers that are “false or plainly unreasonable” themselves. Both Bori and Perkins emphasize that generative AI tools have no understanding of the world or the data they train on. “Philosopherically speaking, it’s bullshit,” Perkins says.
The use of content generated by online news sites is steadily increasing, but this is detrimentally increasing the potential for misinformation being disseminated, and if companies are wrongly mentioned in these hallucinations. are at risk of legal action.
When talking about whether AI will be able to validate its own answers, Bori asks us to consider the methods of scientists.
“It starts with coming up with a plausible hypothesis,” Bori begins. “Then actual experiments confirm or invalidate those hypotheses, and the cycle continues.”
Bori remains skeptical about LLM’s potential to become anything more than a “limited memory machine.” Without a proper roadmap towards generative AI that produces plausible answers every time, generative AI faces another trend of past investment trends.
