Meta and OpenAI expand chatbot voice capabilities, IBM faces hurdles in AI-driven workforce transformation and California mandates one-click subscription cancellations
Welcome to Hashtag Trending, today’s top tech news. I’m your host, Jim Love, let’s get into it.
Both Meta and OpenAI are advancing their AI chatbots with new voice features, intensifying the competition in conversational AI.
Meta’s AI assistant will debut speech capabilities this week, featuring voices from celebrities like Judi Dench and John Cena.
Meanwhile, OpenAI is broadening access to its Advanced Voice Mode for ChatGPT, which offers more natural speech interactions.
There are already a large number of voice based systems on the market, moving rapidly into areas like customer service and even telemarketing. But these new voice chat systems take voice response to a new level. ChatGPT’s voice is scarily close to a human’s. Outside of some minor delays, it can carry on a complete conversation. It can detect and show emotion. It can be interrupted like a human in a conversation.
We haven’t yet tried Meta’s new voice but we expect that it too will be a quantum leap forward.
Despite the breakthrough quality of these voice systems, we can expect some early issues and some work before they are “ready for prime time.” We’ve been testing ChatGPT’s voice, and though impressive, it frequently drops calls or is unavailable. That might be just a byproduct of the early interest.
But even what we’ve seen so far, has been more than impressive. The fluidity of conversation makes it seem human. And – this is a nuance that I picked up. When I asked for some advice on how to use this, the chatbot knew who I was. Not just by name, but it knew what I had used ChatGPT for in past and made specific suggestions based on how I had used it in the past.
While still in early stages, these developments signify a critical shift in AI technology, moving beyond text to more engaging voice-based interactions.
Once the growing pains are under control, expansion of voice AI capabilities could have significant implications for various industries, potentially reshaping how businesses interact with customers and how users engage with technology. And even after these are stable and commercially reliable, there will be new challenges in areas such as voice security, privacy, and the ethical use of AI-generated voices.
So while IT professionals may regard these systems as in their early phases, it would be advisable to have a pilot going to be able to answer when senior management asks when these can be implemented, you can demonstrate what is possible and what is not.
IT leaders need to be thinking about how these advancements will impact their organisations’ AI strategies and user experience designs. They may need to reassess their voice technology infrastructure and explore potential applications and the benefits and risks associated with more sophisticated AI voice assistants.
Sources include: Various
IBM faces hurdles in AI-driven workforce transformation
But while ChatGPT has had two releases that we can say without hyperbole are – ground breaking – it’s o1 release has really brought AI closer to being able to reason – it can solve logic and reasoning problems and mathematical tests that formerly generative AI would have failed. As we noted, it has brought voice conversations that are almost indistinguishable from human conversation including remembering the person it is talking to and reacting to their emotions.
But while OpenAI is basking in its success, one of the early pioneers in AI seems to have lost its way. It’s hard to believe that IBM’s Watson, which shocked the world when it beat the world champions in Jeopardy is now, almost a decade later, fallen far behind others in the industry, at least according to a report from The Register.
According to this story, which featured first hand accounts with three anonymous IBM employees, there is a stark contrast between IBM’s public claims and its internal realities.
IBM had announced ambitious plan to replace thousands of employees with artificial intelligence. But that plan is reported to be encountering significant challenges, and internal realities which make it hard to understand how they will reach their goals.
One employee, a technical specialist, highlighted Watson’s struggles with basic tasks. They stated, “It can’t even do a decent job of creating a PowerPoint. The content is often gibberish or completely irrelevant.” This raises questions about Watson’s readiness for more complex responsibilities.
A second source, working in client relations, expressed concerns about the potential impact on customer trust. They noted, “There’s a huge gap between what IBM says Watson can do and what it actually does in practice. If we start relying on AI for client interactions and it fails, we risk damaging long-standing relationships.”
The third insider, a senior consultant, emphasized the irreplaceable human element in their role. “My job requires understanding nuanced client needs, navigating complex organizational dynamics, and applying industry-specific knowledge. I don’t see how AI can replicate that anytime soon,” they explained.
Even in areas where one would expect IBM to excel, like code generation and testing, employees complained that their internal systems were far inferior to Github’s copilot.
These accounts suggest that IBM’s goal of replacing 7,800 jobs with AI by 2026 may be overly optimistic. The company faces not only technical hurdles but also potential resistance from employees who doubt AI’s current capabilities. Some of the employees questioned felt that the company would still execute on its layoff plans and replace higher priced employees with younger, lower paid workers. Indeed, one referred to the six month cycle of layoffs that was already in place. But this could have devastating impacts on morale and, without proper support, replacing seasoned veterans with inexperienced staff could prove disastrous for a company that has so many high profile enterprise level accounts and who markets themselves as providing expertise and thought leadership.
If these reports are even partially true, we could be seeing one of the first giants in industry to not fall behind because of the inability to develop and leverage, AI, but to have placed their bets on an AI strategy that they can’t achieve.
You can’t count IBM out because of one report, but it is a wakeup call for organisations to ensure that they have a realistic AI strategy.
Sources include: The Register
California mandates one-click subscription cancellations
California has passed a new law requiring companies to offer one-click subscription cancellations if they provide one-click sign-ups. Set to take effect on January 1, 2025, this legislation aims to simplify the cancellation process for recurring subscriptions like streaming services, gym memberships, and digital newspapers.
The new law, Assembly Bill 2863, closes a loophole in existing regulations. While companies were previously required to offer online cancellation options, some made the process unnecessarily complex with multiple steps and opt-out pages. Now, if a service offers one-click purchases, it must provide equally simple cancellation methods.
This change will significantly impact digital product providers, including major players like Roku, Amazon, and Apple, as well as various software developers and content providers. Companies will need to streamline their cancellation processes to comply with the new requirements.
While it doesn’t solve the total problem that many have had to sort out – figuring out from those cryptic notations on credit card statements what a service is and where it came from, it is a start. It’s just our editorial opinion, but we believe that consumers and companies who use subscription services would be better served if specific notice of renewal and a requirement for reauthorization were required. Even if it’s one click, it’s still not effective who and where to cancel are hidden by cryptic billings or in some extensive menu tree. If a service is valuable, people will renew.
For IT professionals, especially those in subscription-based services, this law sets a new standard for user experience and customer retention strategies. It may necessitate updates to billing systems and user interfaces to ensure compliance.
As California often leads in consumer protection legislation, tech leaders should anticipate similar laws potentially spreading to other states and even into Canada. This could require a broader overhaul of subscription management systems for companies operating across multiple jurisdictions.
Sources include: The Desk
That’s our show for today. I’m your host Jim Love, thanks for listening and have a Thrilling Thursday.
