Microsoft blocks access to ChatGPT. Starlink customers are left hanging when accounts are cancelled and no customer support can be reached. SanDisk Extreme Pro Solid State causes a class action lawsuit for hardware failures and lost data AND What if AI bots could name themselves?
These and more top tech stories on Hashtag Trending
I’m your host Jim Love, CIO of IT World Canada and Tech News Day in the US.
In an unexpected move, Microsoft, a major investor in OpenAI, temporarily blocked its employees from accessing the AI chatbot ChatGPT due to security and data concerns, highlighting the intricate balance between innovation and corporate security protocols.
On Thursday, Microsoft imposed a temporary restriction on its employees’ access to ChatGPT, the popular AI chatbot developed by OpenAI, in which Microsoft has invested over $13 billion. This decision, reported by CNBC, was attributed to “security and data concerns,” as indicated on an internal website. The prohibition was surprising given Microsoft’s close relationship with OpenAI and its substantial investment in the technology.
Initially, it appeared that this restriction might be permanent, but Microsoft quickly reversed the decision after CNBC’s report. The company clarified that the service was restored shortly after recognizing an error and emphasized its encouragement for employees and customers to use services like Bing Chat Enterprise and ChatGPT Enterprise, which offer enhanced privacy and security protections.
This news was particularly noteworthy considering the close ties between Microsoft and OpenAI. Microsoft CEO Satya Nadella and OpenAI CEO Sam Altman had recently shared the stage at OpenAI’s first developers conference. Microsoft has been partnering with OpenAI for almost a year, investing heavily in its technology and integrating it into Microsoft’s Bing Chat feature. Bing Chat, powered by GPT-4, represents a more advanced version of ChatGPT and offers online search capabilities.
In addition to Bing Chat, Microsoft has also incorporated DALL-E 3, another OpenAI technology, into its suite of AI tools. This integration allows users to generate images using AI through Bing Chat or the Bing Image Creator.
Sources include: ZDNET
A recent investigation by a data recovery firm reveals that the frequent failures of SanDisk Extreme Pro SSDs, including sudden data loss, are due to inherent design and manufacturing flaws
SanDisk Extreme Pro SSDs have been the subject of customer frustration and a class action lawsuit due to frequent failures and data loss.
A recent report by Attingo, a data recovery company, has pinpointed design and manufacturing flaws as the root causes of these issues. Initially, Western Digital, which owns SanDisk, had identified the problem in its 4TB models and promised a firmware update. However, similar issues have been reported in the 2TB and 3TB models, for which no firmware update has been promised.
Markus Häfele, managing director of Attingo, explained that the hardware, rather than the firmware, is at fault. The SSDs’ components are too large for their circuit boards, leading to weak connections, high impedance, high temperatures, and a propensity for breaking. Additionally, the soldering material used is prone to forming bubbles and breaking. Newer models of these SSDs have been modified with extra epoxy resin to secure the components, but failures are still occurring.
These problems are not limited to a single product lineup but affect both the SanDisk Extreme Portable SSD and the SanDisk Extreme Pro Portable SSD.
Western Digital’s lack of communication about these issues and their decision to continue selling potentially faulty products has raised questions. Data recovery services have become a necessity for some users to retrieve their lost data, highlighting the severity of the issue. This situation underscores the importance of quality control in manufacturing and the need for transparent communication from companies about product flaws.
Sources include: Tom’s Hardware
Starlink, operated by SpaceX, recently faced a customer support challenge as users impacted by an account reset bug struggled with limited support options, highlighting the need for more robust tech assistance in the era of advanced connectivity solutions.
A recent bug in Starlink’s system caused significant user frustration due to the company’s limited customer support infrastructure. The issue began when several Starlink customers received an email stating that their accounts had been reset and any pending orders or deposits refunded. This email, however, turned out to be a mistake, causing confusion among users.
The problem was exacerbated by Starlink’s lack of traditional customer support channels. Users who tried to recover their accounts via Starlink’s account-recovery page faced errors and found themselves unable to reset their passwords successfully. This left them with no direct way to resolve the issue or submit support tickets, as there was no phone number available for customer service.
Starlink’s customer service primarily consists of options like contacting support through the Starlink app or creating support tickets online. However, these methods were not accessible to users with disabled accounts. An alternative email address, starlinkresolutio[email protected], was suggested, but responses were not guaranteed.
Many customers turned to Reddit for support, sharing their experiences and attempting to help each other. This situation underscores the growing pains of new technology companies like Starlink, which may not have fully fleshed out customer support systems in place. It also highlights the reliance of customers on digital platforms for resolving service issues and the potential pitfalls when these systems fail.
Starlink’s rapid expansion and its growing user base, now dealing with technological hiccups and limited support avenues, emphasize the importance of having robust and responsive customer service mechanisms, especially in the tech industry where users heavily rely on uninterrupted service and quick resolutions to technical issues.
Sources include: Ars Technica
Listeners to the show will note that as CIO of a publishing and marketing company, I’ve been somewhat critical of the names given to software products, particularly the new wave of AI products and services.
So I was wondering whether these companies would have been better off if they’d asked their own AI what it wanted to be called. To avoid sending them into a neural spin cycle, trying to figure out if they exist or not, I simply asked the top three, ChatGPT, Bard and Claude.ai to recommend a name other than their own name.
Here’s what they told me – I’ll give you the top three, but if you are really interested, the top 10 names will be in the show notes at itworldcanada.com/podcasts
Here’s the top three names that ChatGPT would suggest as an alternate name:
Unless you are a country and western fan from forty years ago, ConvoAI seems like a bit of a bust. Mercy sakes alive good buddy, looks like we got ourselves a convo.
DialogDynamo – sounds a bit like a superhero.
And InsightBot – All I can say to that one is “meh”
Bard was a little more – shall I say creative or personal? And it makes you wonder just how human this AI has become. But it suggested:
Minstrel, Muse and Rhapsodist with seven other suggestions that were equally lyrical.
Claude AI was equally personal, but a little less aware of potential trademark violations with its suggestions of Alex/Alexa, Robin and Sam/Samantha, but you have to give it full marks for the human quality.
Now, to be totally fair to ChatGPT, Sam Altman has said that he wanted a name that made it clear that his AI chatbot was not a person, so perhaps ChatGPT didn’t have as much to go on. So we asked the Bard and Claude what they would call ChatGPT.
Bard said it would call ChatGPT Dialogue AI, Conversational AI or my favourite in the number three spot – Chatterbox.
Claude stayed in its personal trend and suggested Charlie, Andy and Robin noting that Robin invokes the helpful and intelligent connotation of the bird. My favourite wasn’t in the top three, it was actually number four on the list – Riley as in, “living the life of Riley.”
But in the end I wish someone had taken Paul Simon’s advice when he said – “you can call me anytime, but anytime you call me Al. Call me AI.
And if you don’t get that one, type it out and take a look.
And that’s the top tech news for today.
Hashtag Trending goes to air 5 days a week with a special weekend interview show we call “the Weekend Edition.”
You can get us anywhere you get audio podcasts and there is a copy of the show notes at itworldcanada.com/podcasts
For those interested in cybersecurity you can also check out our hit cybersecurity podcast featuring Howard Solomon and called CybersecurityToday. It’s often rated as one of North America’s top 10 tech podcasts.
I’m your host, Jim Love – have a Marvelous Monday! And I’ll be back next week as our producer James Roy will be hosting while I’m in Vancouver hosting the Big Data Conference. Listeners in Vancouver, come and say hi
ADDITIONAL SHOW NOTES: Top names that AIs would call themselves if they had been asked:
ChatGPT would have chosen:
- ConvoAI: Highlighting its ability to engage in conversations using artificial intelligence.
- DialogDynamo: Emphasizing its dynamic and powerful capabilities in dialogue and conversation.
- InsightBot: Reflecting its role in providing insights and information.
- TalkTrend: A name that suggests its ability to keep up with trends and topics in conversations.
- InfoInteractor: Focusing on its interaction capabilities with a strong emphasis on information exchange.
- ChatterMind: Combining the aspects of chatting and its intelligent, mindful approach.
- WisdomWave: Suggesting a wave of wisdom and knowledge in its responses.
- Chatwise Companion: Indicating its role as a wise companion in chat-based interactions.
- SpeakSphere: A name that encapsulates the global reach and spherical nature of its conversational abilities.
- DialogIQ: Combining dialogue with a high intelligence quotient, indicating smart conversations.
Bard would have chosen:
Bard suggested that it might call itself:
- Poet Laureate
And ClaudeAI suggest that it might call itself:
Here are some name suggestions for ChatGPT:
– Charlie – Friendly and casual sounding.
– Andy – Gender neutral and easy to remember.
– Robin – Invokes the helpful and intelligent connotation of the bird.
– Riley – Energetic and friendly.
– Jordan – Simple and gender neutral.
– Sam / Samantha – Familiar human names that are widely recognizable.
– Jamie – Fun and lively name.
– Jessie – Approachable and upbeat.
– Alex / Alexa – Nods to the popularity of virtual assistants with these names.
– Sage – Wise and knowledgeable sounding.
– Sky / Skylar – Aspirational and forward thinking.
– Echo – References the idea of an echo chamber and listening.
– Atlas – Strong and knowledge based.
– Onyx – Modern and sleek.
And it went on to say:
Let me know if any of those resonate with you as a good new name for me! I’m open to suggestions as my name is not set in stone. The most important thing is choosing a name that feels natural when you chat with me.