Tech giants like Apple, Google, and Meta are deeply involved in the AI race, with Apple reportedly testing an internal chatbot named Veritas to enhance its Siri assistant. Veritas, which functions similarly to ChatGPT and Gemini, allows employees to test features like searching personal data and performing in-app actions. While Veritas is not slated for public release, Apple is also developing a revamped Siri, codenamed Linwood, which will integrate external models with its own technology for web searches and on-device data. Google's Gemini is also being considered for Apple's AI-powered search capabilities. This intense AI development mirrors the dot-com bubble era, characterized by massive investment and high valuations, though major AI players are currently generating substantial revenue. However, a significant challenge persists, as a recent MIT study reveals that 95% of companies are not yet seeing returns on their AI investments, and generative AI itself might not achieve the economic powerhouse status once predicted, especially with the rise of free, open-source models like Meta's Llama. The potential for AI 'hallucinations' also presents risks, as seen in a legal case where an attorney faced sanctions for submitting fake legal quotes generated by ChatGPT. In programming, AI-generated code is sparking debate about quality and maintainability, with some experts concerned about an influx of mediocre code. The burgeoning AI sector is, however, revitalizing San Francisco's office market, which has been struggling with high vacancy rates. Meanwhile, some suggest that the term 'artificial intelligence' itself might be misleading, proposing 'aggregated intelligence' as a more accurate descriptor for current AI's ability to synthesize information rather than create novel ideas.
Key Takeaways
- Apple is testing an internal AI chatbot named Veritas to improve Siri, with features similar to ChatGPT and Gemini.
- Veritas is not planned for public release, but Apple is working on a revamped Siri (Linwood) that will combine external and internal AI models.
- Apple may use Google's Gemini for its AI-powered search capabilities.
- The current AI boom shows parallels to the dot-com bubble, with high investment and valuations, though leading AI companies are generating revenue.
- A study found that 95% of organizations are not seeing returns on their generative AI investments.
- Generative AI might not become the economic powerhouse predicted, especially with the availability of free, open-source models like Meta's Llama.
- AI 'hallucinations' pose risks, as demonstrated by an attorney facing sanctions for using fake legal quotes generated by ChatGPT.
- Experts are debating the quality and maintainability of AI-generated code in the Go programming community.
- The AI sector's growth is helping to revive San Francisco's struggling office market.
- Some propose 'aggregated intelligence' as a more accurate term for current AI, which synthesizes information rather than creating new ideas.
Apple tests new Siri AI with internal chatbot Veritas
Apple is testing upgrades for its Siri AI assistant using an internal chatbot named Veritas. This tool allows employees to test new features like searching personal data and performing in-app actions. Veritas functions similarly to chatbots like ChatGPT and Gemini, enabling back-and-forth conversations. However, Apple has no plans to release Veritas to the public. Instead, the company is likely to use Google's Gemini for its AI-powered search capabilities.
Apple's internal AI chatbot Veritas could be public
Apple is testing a new internal chatbot called Veritas to improve its Siri AI. This tool allows employees to test features like searching personal data and performing in-app actions, similar to ChatGPT. While Apple has not announced plans for a public release, some believe launching Veritas as a standalone app could increase interest in their AI advancements. The company is also working on a revamped Siri, code-named Linwood, which will combine external models with Apple's own technology for web search and on-device data.
Is 'artificial intelligence' the right term?
A letter writer suggests that 'artificial intelligence' is a misleading term for current AI technology. They propose 'aggregated intelligence' or 'aggregated information' as more accurate descriptions. The writer believes AI currently synthesizes information rather than creating truly new ideas. They suggest that only when AI can develop autonomous solutions might 'advanced intelligence' be a fitting term.
Cultural snobbery may be needed in the age of AI content
The article questions whether cultural snobbery should return as AI-generated content and bland entertainment become more common. It notes that in the past, judging art by its sophistication was seen as elitist, but now, with the rise of algorithmic content, consumers might need to be more discerning. The author suggests that while poptimism embraced all forms of culture, it may have led to indiscriminate celebration. This shift could be important as AI potentially floods culture with less meaningful content.
AI 'hallucinations' pose risks in legal settings
A University of San Diego law professor, David McGowan, discusses the dangers of AI 'hallucinations,' where AI tools generate incorrect information. He highlights a case where an attorney was fined $10,000 for submitting a filing with fake legal quotes generated by ChatGPT. McGowan warns that lawyers must double-check AI-generated content to avoid sanctions. He also expresses concern about people using AI for legal documents without understanding the risks, though he acknowledges AI can be beneficial if used correctly.
Go programmers debate AI-generated code quality
Go programming experts are discussing the impact of AI-generated code on the language's popularity and quality. While AI tools make it easier to build software, programmers like Jonathan St. Pierre express concern about maintaining low-quality AI code. They note that the ease of AI code generation leads to many mediocre options, making it difficult for consultants. Despite these challenges, some see AI as a way to introduce more people to programming.
AI boom echoes dot-com bubble fears
The current AI boom shows similarities to the dot-com bubble of the late 1990s, with high company valuations based on potential rather than profit. Like the internet boom, AI investment is massive, with tech giants pouring billions into infrastructure. However, unlike many dot-com companies, major AI players are generating significant revenue. The key lesson from the dot-com crash is that even transformative technologies must eventually justify their valuations with real business results.
AI demand boosts San Francisco office market
The demand for artificial intelligence is revitalizing San Francisco's struggling office market, which has faced high vacancy rates. A Wells Fargo Economics report indicates that the AI sector's growth is turning the market around. This surge in interest from AI companies is creating new opportunities and demand for commercial real estate in the city.
Generative AI could be worthless, and that's okay
The article explores the possibility that generative AI might not become the economic powerhouse predicted, and suggests this outcome could be positive. Current AI tools are costly to run and may not generate sufficient profits through advertising alone. Issues like copyright infringement and the expense of legal battles could make AI models liabilities. The rise of free, open-source AI models like Meta's Llama also challenges the high valuations of commercial AI firms, suggesting a future where AI is accessible and functional, even if not highly profitable.
Most companies see no return on AI investments
A recent MIT study found that 95% of organizations are not seeing any return on their substantial investments in generative AI. Despite widespread adoption, companies are struggling to achieve positive financial results from their AI initiatives. This suggests a significant gap between the investment in AI technology and its practical, profitable application in businesses.
Sources
- Apple’s ‘Veritas’ chatbot is reportedly an employee-only test of Siri’s AI upgrades
- Apple’s ChatGPT-Style Chatbot App Deserves a Public Release
- What word should go after
- In an era of AI slop and mid TV, is it time for cultural snobbery to make a comeback?
- Local professor weighs in on AI "hallucinations" in court
- Go Experts: 'I Don't Want to Maintain AI-Generated Code'
- Everyone's wondering if, and when, the AI bubble will pop. Here's what went down 25 years ago that ultimately burst the dot-com boom
- AI needs more humans: AI company demand jolts S.F. office market
- Generative AI might end up being worthless — and that could be a good thing
- 95% companies have seen zero return on their AI investments and the reason is ..., says study - The Times of India