Today on the AI Daily Brief, Menlo Ventures published its 2024 State of Enterprise AI report. Before that in the headlines, OpenAI had been hiring key developers of Chrome away from Google to develop their own browser. The AI Daily Brief has a daily podcast and video about the most important news and discussions in AI. To join the conversation, follow the Discord link in our show notes.
Welcome back to the AI Daily Brief Headlines edition, all the daily AI news you need in around five minutes. Our first story today, OpenAI is apparently considering taking on Google directly by launching their own browser. According to a scoop from the information, OpenAI have been hiring key developers of Chrome away from Google to develop their own browser. This comes just weeks after the launch of chat GPT search interface, and already OpenAI has been pitching the product for integration with other websites and apps.
The information sources claim that Conde Nast, Redfin, Eventbrite, and Priceline are all involved in discussions. The pitch is reportedly that OpenAI could help transform customer interactions, allowing website visits to include natural language search assistance trained on the site's contents. By way of example, a clothing store could include an assistant to help with styling question based on the store's particular inventory. The company is reportedly also in early talks with Samsung to integrate AI features onto their headsets.
another assault on Google's dominance with its Android platform. All of this adds up to what looks like a big business development push with OpenAI, trying to model a Google-style ecosystem rather than just supplying AI models. Then again, it all feels very early. Sources say the product is still facing technical challenges around data security, privacy, extension compatibility, basically everything. But it still has people excited.
veteran AI product guru, Alan Masal-Glu wrote, once OpenAI brings out a browser, it's good game for everyone. They will eat everything. I'm pretty confident to say that they will easily be able to capture a huge amount of consumer market across a wide range of horizontals and verticals. When someone asked, what could they do with a browser that current companies can't, Alam said, it's not about capabilities itself, but they already have their branding positioned extremely well. If you ask any normie what AI is, they will say back OpenAI.
It is a long-term trope that ambitious tech companies at some point always set their sights on a browser, usually to be disappointed. Nick Wingfield, for example, writes, you haven't lived as a major tech company until you roll your own web browser. But this competition between OpenAI and Google obviously seems more opportune. Of course, it's not just OpenAI's technology that's causing Google struggle. Google is in the midst of a huge antitrust case that could see them forced to unwind Chrome.
Additionally, we're getting information that according to the most recent proposal from the DOJ, antitrust sanctions against Google could have an even more profound impact on the AI industry as well. As part of the resolution of the case, which is about Google's search monopoly, the Justice Department proposed restrictions on investment, which would force Google to unwind their partnership with anthropic. The restrictions would prevent Google from acquiring, investing in, or collaborating with any company that controls where consumers search for information, including query-based AI products.
Bloomberg sources familiar with the DOJ's thinking said that this proposal was specifically intended to apply to Google's investment in enthropic. Google committed to investing $2 billion in the company last October with $500 million being paid up front. The deal was structured to keep the tech giant at arm's length receiving only non-voting shares and mere consultation rights over major business decisions.
In a blog post, Google hit back at what they call an extreme proposal. They wrote that the proposal would, quote, chill our investment in artificial intelligence, perhaps the most important innovation in our time where Google plays a leading role. They went on claiming that, quote, the DOJ's approach would result in unprecedented government overreach that would harm American consumers, developers, and small businesses and jeopardize global economic and technological leadership at precisely the moment it's needed most.
Interestingly, the revelation comes just one day after a UK regulator discontinued their investigation into the Google Anthropic Partnership, finding that it was not a de facto merger. The DOJ proposal is not final yet. It would need to be approved by a judge before it can be enforced. Staying on our big tech theme, Apple is apparently stepping up development of their AI-powered version of Siri. According to Bloomberg sources, Apple is racing to develop a more conversational version of Siri to keep up with OpenAI and other rival voice-enabled chatbots.
By far, this is the most desired thing when it comes to Apple intelligence and whatever that is going to be ultimately. And while the project to overhaul Siri has been rumored for months, amid the sort of nonchalant rollout of Apple intelligence features, this reporting takes on a new tone implying that Apple has finally recognized how important it is to catch up on AI.
As for features, the product which has been codenamed LLM Siri, very mysterious guys, will be capable of having more natural and conversational voice interactions. The Assistant will also have expanded ability to use App Intents to interact with third-party apps. Apple has already been using LLM integrations with Siri in their latest iOS beta. When appropriate, Siri can hand off queries to chat GPT for a more detailed answer. This project seems to be an attempt to bring that feature in-house, driven by Apple's proprietary models.
The decision to commit resources seems to be a relatively recent one, with Apple hiring over the last month for a team that will quote, create groundbreaking conversational assistant technologies. According to Bloomberg reporting, the new series will be unveiled next year with plans to launch in the spring of 2026.
The question is whether they're already too late. Investor Marcelo Lima wrote, OMG, Apple is such a snail. iOS 18.1 has no real intelligence to speak of and Siri remains just as dumb. Meanwhile, the world is zipping by them fast. Spring 2026 is like a decade from now.
And it certainly sort of feels like that from where I'm sitting as well. But at the same time, it's probably a good idea never to entirely count out to Apple. But so far, it does seem like they really have been a few steps and potentially a few disastrous steps behind when it comes to generative AI. That, however, is going to do it for today's AI Daily Brief Headlines edition. Next up, the main episode.
Today's episode is brought to you by Vanta. Whether you're starting or scaling your company's security program, demonstrating top-notch security practices and establishing trust is more important than ever. Vanta automates compliance for ISO 27001, SOC 2, GDPR, and leading AI frameworks like ISO 42001 and NIST AI Risk Management Framework, saving you time and money while helping you build customer trust.
Plus, you can streamline security reviews by automating questionnaires and demonstrating your security posture with a customer-facing trust center all powered by Vanta AI. Over 8,000 global companies like Langchain, Leela AI, and Factory AI use Vanta to demonstrate AI trust and prove security in real time. Learn more at vanta.com slash NLW. That's vanta.com slash NLW.
Today's episode is brought to you as always by Super Intelligent. Have you ever wanted an AI daily brief but totally focused on how AI relates to your company? Is your company struggling with AI adoption either because you're getting stalled figuring out what use cases will drive value or because the AI transformation that is happening is siloed at individual teams, departments and employees and not able to change the company as a whole?
Super Intelligent has developed a new custom internal podcast product that inspires your teams by sharing the best AI use cases from inside and outside your company. Think of it as an AI daily brief, but just for your company's AI use cases. If you'd like to learn more, go to bsuper.ai slash partner and fill out the information request form. I am really excited about this product. So I will personally get right back to you. Again, that's bsuper.ai slash partner.
Welcome back to the AI Daily Brief. Today we are doing a fun one, Menlo Ventures, which is a Silicon Valley firm that is very, very active in the AI space, has published its 2024 state of enterprise AI report. The results come from surveys with 600 or more key buyers across US enterprises, and there are some really interesting findings.
Apologies in advance that this will be a little less visually interesting than our normal show, but what we're going to do is go through their major conclusions and considerations, and I'll provide my commentary based on what we're seeing in the market as well.
One of their first conclusions is that enterprise AI is moving from pilot to production. The big stat here is that AI spending was up more than 600% from the 2.3 billion that was spent in 2023 all the way to 13.8 billion this year. Menlo calls that a, quote, clear signal that enterprises are shifting from experimentation to execution.
They also note that this comes with significant organizational optimism. 72% of decision makers, they say, anticipate broader adoption of generative AI tools in the near future. And yet, it is still very clearly early. More than a third of Menlo's respondents said that they didn't have a clear vision for how generative AI will be implemented across their organization.
So a couple follow-ups from me. First, it is definitely the case that, one, there is increased spending and increased activity around AI, and that, two, there's increased optimism in a sense of broader adoption coming. My only quibble here is with the framing of from pilots to production. Generative AI is very much not a technology shift that's going to fit comfortably into phases like that.
What I mean is that what we're seeing across the enterprise is that even when pilots are successful and they move into broader institutional adoption, there's a whole additional set of pilots that are just getting started and are earlier in their journey. And what this reflects is that generative AI marks a shift in how frequently organizations are going to need to change. It's not a one-time transition from pre-AI to post-AI. It's a new and ongoing process.
Many organizations even that have put significant AI initiatives into production still feel behind in other areas. And that's of course because generative AI is so wide reaching in its implications. Still to the extent that the conclusion here is that 2024 was a year of more, absolutely, absolutely true.
The next stat, though, tells a really important part of the story as well. Menlo found that 60% of enterprise-generated AI investments come from innovation budgets. That obviously reflects the fact that we're still early. However, that means that 40% of spending is sourced from more permanent budgets. And of that, they found 58% was redirected from existing allocations. The conclusion they come to is that that demonstrates a growing commitment to AI transformation, and I think that's correct.
As an aside, one of the additional things that we've seen is that the sector of the organization that was dedicated to innovation, while sometimes previously kind of toothless and under-resourced, has become a much more significant player in the generative AI era. We're finding that the most successful organizations have reconstituted and reimagined their innovation organizations, not as a silo separate from the rest of the company, but instead as a coordination function that touches every other part of the company.
that understands the AI implications and issues for legal, compliance, business units, basically everyone you can imagine. The innovation part of the organization is getting upscaled alongside the proportional value of artificial intelligence. The next thing Menlo focuses on is what organizations are actually buying. And the biggest finding here is that organizations are no longer just spending on foundation models, but are actually making purchases at the application layer. Whereas in 2023, Menlo found enterprises spending 600 million on AI applications,
This year, that number was $4.6 billion and an 8x increase. Menlo also writes, companies aren't just spending more, they're thinking bigger.
On average, organizations have identified 10 potential use cases for AI signaling broad and ambitious goals. Nearly a quarter of those are prioritized for near-term implementation. And this, of course, gets at the idea of growing sophistication. The increasing spend on applications reflects the fact that a lot of enterprises are now a couple layers deep in their AI journey. They've done some pilots, found some things that work, and are doubling down on new transformative workflows that improve or innovate what they do.
But what are they doing with AI? Well, that's the next thing that Menlo explores. They found in order that the most adopted use cases in the enterprise were code co-pilots with 51% adoption, support chatbots with 31% adoption, enterprise search retrieval and data extraction at between 27 and 28% adoption, and meeting summarization in fifth at 24% adoption. Interestingly, copywriting came in sixth at 21%, and image generation seventh at 20%.
The reason that's interesting is that obviously writing an image generation are sort of the most dominant consumer use cases, so you're also seeing here, I think, more sophistication on what constitutes an enterprise versus a consumer use case. This also looks to me like a story of starting by starting. A lot of the use cases that are most embraced already are the one-to-one replacements for old processes with something new and better.
Next up, one of the big questions that many people are asking is to what extent enterprises are going to build their own custom solutions versus just buy from external vendors?
Menlo saw a major shift between 2023 and 2024 on that front. In 2023, they found 80% of enterprises relying on third-party Gen AIs software, while in 2024, 47% of solutions were developed in-house, as opposed to 53% source from vendors. Those numbers are dramatic, but we're certainly seeing something similar, at least in the sense that a couple things are happening.
First of all, a lot of the higher order use cases for generative AI rely on interaction with the company's data. And so there's already an inherent bias and push towards an in-house solution that has fewer security concerns. At the same time, as companies move towards more specialized use cases,
Vertical and functional AI applications are just starting to come online. And so I wouldn't be surprised if we're in a period for some time where before vertical and functional AI applications reach dominance, a lot of organizations experiment with spinning up their own in-house versions of them first. In other words, I wouldn't be surprised if over a four or five year time period,
We see first the shift that we're in right now from third party vendors to in-house solutions and then back a couple years later towards more third party solutions as they get better and more specific. Another really interesting finding is that AI isn't just hitting one area, it is really hitting every different area.
IT commands 22% of the spend, product and engineering commands 19%, customer support 9%, sales 8%, data science 8%, marketing 7%, human resources 7%, accounting and finance 7%, design 6%, legal 3%, and everything else still another 6%. This is really truly a org-wide transformation, which of course is part of what makes it such a challenge for companies that haven't had to think that broadly and comprehensively about a transformation process, maybe ever.
There is a lot more in this survey, but we'll close on just one question that obviously takes a huge amount of time and attention. What is really the state of agents? It's clear from this survey that agentic architectures are still mostly in the realm of the future. However, Menlo did note that from 0% in 2023, agentic architectures now make up 12% of implementations.
I would expect that this number will go up wildly in the next year, but we will of course have to wait for it next year's survey to know for sure. For now though, that is going to do it for the AI Daily Brief. Until next time, peace.
