Imagine a world without Facebook. For just half an hour. A group of AI leaders get together but don’t seem to invite Elon Musk. Amazon takes data centres nuclear. A new competitor for ChatGPT and Google. And public trust in AI is declining.
 |
 |
 |
All this and more on the “just breath, Facebook will be back” edition of Hashtag Trending. I’m your host, Jim Love, CIO of IT World Canada and TechNewsDay in the US.
We heard the collective scream as hundreds of thousands of users of Meta’s social media sites – Facebook, Instagram and Threads were locked out of their accounts early this morning. The details are still coming in as we record this show for tomorrow morning, but here’s what we know.
Between about 10 am Eastern time and noon, a large number of users were locked out of their accounts and unable to login to Facebook, Instagram and Threads.
Initial reports posted on Cyberint.com claimed that three threat actor groups – Skynet, Godzilla, and Anonymous Sudan – claimed that they had shut down Facebook, Instagram and Threads.
The numbers affected were in the hundreds of thousands. The site Downdetector logged over 560,000 US Facebook users unable to get on the site and about 160,000 in Canada. The proportions were the same on Instagram and Threads.
But not everyone was affected. While I was viewing comments from users in Toronto and Ottawa on Downdetector, I was able to login into Facebook from my house which is a two hour drive to Toronto.
Looking at other sites including Twitter and others, there was a similar spike at 11 am although nothing like the number of users affected on Meta’s sites. These were larger than normal but only in the hundreds.
I’m sure that there will be more on this. Check out ITWorldCanada.com for an update this morning and watch out for more from Howard Solomon, host of CyberSecurityToday.
I’m sure this will be part of the discussion today and maybe even in days to come.
Sources: Downdetector, Cyberint.com, ITWorldCanada.com
As Meta’s users were being locked out, Elon Musk took the time to make a clear statement about how attacks on any sites are unacceptable. Sorry, just kidding.
Elon took to X to mock the outage, with a meme and in a separate post, he also added, “if you’re reading this post, it’s because our servers are working.”
Always a classy guy.
Almost as if in response to Elon Musk’s lawsuit against OpenAI, a new coalition of AI companies that has published an open letter sponsored by SV Angel, a thirty year old investment firm headed by founder Ron Conway. Their opening to the letter says, “we invest in and support founders in AI who are committed to building the next generation of AI technology to move humanity forward.
The letter says, “We, the undersigned, already are experiencing the benefits from AI, and are committed to building AI that will contribute to a better future for humanity – please join us!”
The letter appears to be open to everyone to sign, but the initial signatures are a virtual “who’s who” in AI – OpenAI, Google, Mistral, Meta, Salesforce, Hugging Face and the list goes on and on, with one noted exception. Elon Musk didn’t appear to be one of the initial signatures. Is that because he doesn’t believe in it?
Or is it that the other kids don’t want to play with him anymore?
Check it out and sign it at openletter.svangel.com
Sources include: svangel.com
Amazon Web Services (AWS) has moved to nuclear-powered data centers to reduce its carbon footprint. A $650 million acquisition deal with Talen Energy grants AWS ownership of a data center complex, situated adjacent to Talen’s Susquehanna nuclear power plant in northeast Pennsylvania.
Initially opened in early 2023 with a 48-megawatt capacity, there are plans to expand Cumulus to 475 megawatts, with potential for further growth with Amazon’s investment.
Talen Energy will receive $350 million upon the deal’s closure, with an additional $300 million to follow based on the completion of development milestones throughout 2024. Talen will continue to supply AWS with direct access to electricity generated by the nuclear power plant, with AWS expecting to have up to 960 megawatts at its disposal in the coming years. If I did my calculation right, that’s enough to power almost a million homes, just to put it into perspective.
This is just one of Amazon’s ongoing efforts to power its data centers with carbon-free and renewable energy sources. Last month, Amazon entered a power purchase agreement with a wind farm in Oregon, and it has explored alternative energy solutions, including natural gas fuel cells, which can be converted to run on hydrogen.
The move into nuclear energy is not unique to Amazon; the data center industry is increasingly considering small nuclear reactors (SMRs) for their modularity and potential cost benefits, despite none being operational yet.
With the advent of AI and the growing need for power for data centres, finding working solutions quickly is critical.
Sources include:
Anthropic, founded by Dario Amodei, former vice president of research at OpenAI, has been a solid number three in the big generative AI players. It originally distinguished itself as a rival to ChatGPT close to equal in ability, but having an advantage of a huge context window, allowed it to take in larger questions than OpenAI.
Claude.ai, their generative AI model has competed in a way almost reminiscent of the old Mac computer compared to the bigger Windows machines. It’s been a favourite of writers and those who love it, love it. There’s a dedicated fan base on places like YouTube who sing its praises.
But the most recent release has set YouTube ablaze. Claude 3
now claims that it outperforms OpenAI and other generative AI rivals.
The announcement, made via a blog post, introduces three foundational models named Opus, Sonnet, and Haiku, each designed with enhanced analysis, forecasting, code generation, and content creation capabilities.
Claude 3 Opus has trotted out a series of scores on tests that claim to show superior results such as the HumanEval coding test, scoring 84.9 per cent, compared to Google’s Gemini Ultra at 74.4 per cent and OpenAI’s GPT-4 at 67 per cent.
There are a series of other measures and they are all impressive. Fair warning, these tests, like human IQ tests have been controversial, but still, for a third-place performer to make a leap like this has made waves.
With Google trying to launch and Claude.ai making a big splash, we can only wonder if OpenAI will push the development of its next big move. We know GPT 5 is in testing and rumours abound about its potential, and a new development called Q*
Sources include: CIO Dive, YouTube
Public trust in artificial intelligence (AI) technology and the companies developing it is declining globally, according to new data from Edelman and shared with Axios.
This trend is evident both in the United States and internationally, with trust levels dropping from 61 per cent to 53 per cent globally over the past five years. In the U.S., trust has seen a significant decrease of 15 percentage points, from 50 per cent to 35 per cent in the same period.
Edelman’s global technology chair, Justin Westcott, emphasizes the importance of trust in the AI era and calls for companies to address the broader implications of AI technology, including its social impact and the protection of personal privacy.
The study also highlights a political divide in trust levels within the U.S., with Democrats showing slightly higher trust in AI companies (38 per cent) compared to independents (25 per cent) and Republicans (24 per cent). Furthermore, the technology sector, which was once the most trusted industry in 90 per cent of the countries Edelman studies, now holds the top spot in only half of those countries.
Interestingly, the acceptance of AI varies significantly between developed and developing countries. In developed nations such as France, Canada, Ireland, the UK, the U.S., Germany, Australia, the Netherlands, and Sweden, there is a strong resistance to the growing use of AI. In contrast, developing markets like Saudi Arabia, India, China, Kenya, Nigeria, and Thailand show a much higher acceptance rate.
Edelman’s research also points to public dissatisfaction with government efforts to regulate AI, with many feeling that current regulations do not meet their expectations. This sentiment underscores the urgent need for regulators to address public concerns about AI more effectively.
Sources include: Axios
That’s our show for today.
Love your comments.
Send us a note at [email protected] or drop us a comment under the show notes at itworldcanada.com/podcasts – look for Hashtag Trending.
Thanks for listening and have a Wonderful Wednesday.
