This week’s most important AI news

An AI-generated image of artificial intelligence in the workplace. Source: Private Media

It’s been a big week for AI.

While this is an overwhelmingly true statement this week, it’s also one that is starting to feel somewhat redundant. Increasingly, every week is a big one for AI. Between big tech companies rolling out new tools, AI-focused startups getting significant cash injections and ethical boundaries being pushed — there is always something going on.

After OpenAI released ChatGPT to the public in November 2022, we made an editorial choice to consistently report on artificial intelligence at SmartCompany. We saw the impact it was going to have on businesses, particularly on the startups and SMEs its our mission to servce.

Things took off quickly. Against the backdrop of the economic downturn, a drop in capital raises (compared to 2021, in any case), startups collapsing and job cut after job cut after job cut — generative AI in particular, grew.

Every major software company raced to get their tools out the door. Microsoft, Google and Meta came to the party swiftly, even if sometimes things didn’t go to plan.

The pursuit of business-related AI stories has taken me to the US three times over the past six months — from grilling Adobe’s vice president of generative AI on copyright and artist protection to observing how Salesforce is using the power of marketing and messaging to rip the fear out of AI.

At one point I even found myself racing across Seattle to Microsoft’s Redmond campus because the tech giant had another round of layoffs on the same day it announced a suite of new AI-powered security features.

Because ultimately, what happens in the world of AI globally has a trickle-down effect. It can, and will have an impact on Australian businesses that are actively utilising this technology, and even those that aren’t by way of changing the world of tech in the long run.

The problem is, new developments in AI are relentless. Every single day there are multitudes of stories, announcements, and controversies. It can be difficult to keep up.

After almost a year of being on this beat, along with general business tech and Australian startup news, I came to the conclusion that there had to be an easier way for people to find these stories.

And that has led us to Neural Notes: a new weekly column where I condense all of the most important AI news of the week for you.

This will be a place where anyone with an interest in AI and business can come to learn about what’s happening in the space, with explanation, contextualisation, and a sprinkle of my signature sass: it definitely won’t just be a link dump.

We’re experimenting with the format and story inclusions for now, so if you have any suggestions of what you’d like to see, let me know!

Let’s get cracking.

Big AI tool announcements from Meta, OpenAI and Getty

The new AI tools news has been coming in thick and fast this week, with Meta leading the pack with a ton of new tools across Instagram, WhatsApp and Messenger.

While a lot of it is aimed at individuals, its new AI Studio is directly aimed at businesses that want to create their own no-code AI chatbots. Its new generative AI image tools also have some solid applications for businesses that utilise Instagram. You can read our full story on this here.

But this wasn’t the only big news for AI image generation this week.

OpenAI announced the latest upgrade to its image generation tool, DALL-E 3. While this is the third iteration of DALL-E, paid ChatGPT users can now use the platform to get images as well as word-based results. The company also revealed plans for voice in the future — with users being able to one day submit prompts via voice and receive a spoken response from the platform.

Getty also announced its foray into AI image generation, backed by AI hardware darling Nvidia’s Picasso LLM. Similar to Midjourney and OpenAI’s DALL-E 3 models, users are able to create an image from a text prompt.

However, Getty is differentiating itself by pushing the fact that Picasso was trained in Getty’s own library which promises “uncapped indemnification” when it comes to intellectual property claims. What Getty is essentially promising here is that the AI images it generates will be commercially safe. And similar to Adobe, it says that it is properly compensating creators.

Speaking of Adobe, it also pushed its image offerings further this week. It announced that Photoshop for web was coming out of a two-year beta. And it’s bringing its Firefly AI tools with it. Massive.

Commercial safety is a huge consideration when it comes to using generative AI for business purposes — particularly as regulation continues to roll out in different jurisdictions across the globe.

What’s particularly interesting about this space is that when it comes to commercial assets created by generative AI, there is still a question mark around copyright protection here in Australia. This is why it’s so important to choose your tools wisely, preferably from an ethical company that can ensure that what you create is legally yours.

AI cash splash

There’s been a few AI-focused raises here in Australia during September, including $15.25 million for Neara and $1.8 million for Eloise has all the details for you.

Over in the US, there has been a lot of VC money being thrown at AI this week alone:

  • US$4 million for Anthropic, led by Amazon. These guys are pegged to be the next OpenAI and this is not their first raise of the year. Back in March, it was included as part of Salesforce Venture’s $369 million AI-focused fund (which has since doubled to over $778 million). This was solidified in May when Anthropic announced its US$450 million series C. This was led by Spark Capital, with participation from Salesforce Ventures, Google, Sound Ventures, Zoom Ventures, Menlo Ventures, and more.
  • OpenAI is reportedly looking to sell shares in order to boost its valuation from US$29 billion to between US$80 and US$90 billion. Watch this space on that one.
  • US$10.6 million for Nexusflow, which focuses on generative AI for cybersecurity tools.
  • US$150 million for AlphaSense, which specialises in utilising AI for market research. This is its Series E round and comes just months after a US$100 million Series D. Huge.
  • US$15 million for Kolena, which actually tests and validates the performance of AI models. This could be significant in the future as the barrier to businesses creating their own AIs lowers.
  • A whopping US$49 million for Kneron: a developer of AI chips for self-driving cars.

Regulation and controversy

One of the biggest issues around generative AI is works being used to train LLMs without permission. It’s happening everywhere — from websites to artwork to books.

And most people who have a problem with this say the same thing — there needs to be an opt-in functionality. It seems we’re still a way off from that, so for now what we are seeing more of is opt-outs.

And that’s what Google has announced this week. It will now benevolently allow online publishers — like us — to deny its Bard and Vertex AI platforms from learning from them. Or, in Google’s words:

“Manage whether their sites help improve Bard and Vertex AI generative APIs.” Uh-huh.

But the biggest controversy of the week has been about the 191,000 books being used to train the Book3 LLM without the consent of authors. This has included extremely well-known international and Australian authors like Stephen King, Margaret Atwood and Tim Winton.

Knowledge about this dataset has been kicking around for a while and has resulted in several lawsuits against the likes of OpenAI and Meta. However, it has come back around this week as more authors discovered they were on this list, particularly comparatively smaller ones with followings across TikTok.

While it is still early days on the regulation front — the law is always behind tech — this is not a good look. Increasingly, consumers want to deal with ethical companies. As more and more tech companies build out their own AIs, people will have a choice.

This highlights why it’s always important to ask not just if you should begin utilising AI, but how you do it. It’s going to be fascinating to see how this plays out and whether these tech giants will be forced to pivot from not just trying to be the quickest to market with their AI offerings, but the ones doing it the right way.

Thanks for reading! Do you have an AI-related tip or story? Let us know for the next edition!

The Post This week’s most important AI news Originally Posted on

Leave a Comment

Scroll to Top