Lead Story – What’s Next for Enterprise AI?Focus on specialization and regulation
Enterprise AI desperately needs accuracy – what does it do and where does it fall short? What are your concerns? Where’s the ROI?
Yes, you might think these are straight-forward questions, but judging by the AI hype festivities in my daily inbox, they’re clearly not. However, Chris moves forward through his AI. Clari CEO says niche is good when it comes to enterprise GPT. As Chris explains, Clari has a skin for this particular game.
Earlier this month, the 10-year-old company launched RevGPT. His new GPT-powered tool queries data stored in Clari’s RevDB database, so users can ask key revenue-related questions in areas such as risk, sales targets, and forecasts and can act accordingly.
One of the benefits of generative AI in the enterprise is small/niche data sets refined (“smarter”?) by iterative user feedback.
Unlike ChatGPT instances, which are trained and referenced with data scraped from the web pre-2021, RevGPT and its companies only have access to company-specific and trusted data. Over time, according to Clari, employees will experience a “flywheel effect” of greater accuracy in their answers, actions and results.
Admittedly, this does not address overall concerns about the far-reaching impact of AI. But it’s that containment/focus that makes this use case possible.
The strength for us is that the use cases are limited. RevGPT is not this generic “ask everything in the world” application. And when you combine that with the unique data we have, we can make powerful predictions and recommendations at every “earnings moment” for every employee where revenue matters.
Other departments and roles have different GPT usage examples. However, I would like to point out that even if the AI is well trained, an accurate recommendation/prediction is not a completed transaction. Data quality/depth/diversity factors, including external trend data, play a big role. As a broader concern, Derek revisits that while the topic of AI regulation is a priority, awareness and education are also priorities. Let’s not repeat the same mistakes.
Governments, the private sector and research institutes are working together to develop several national campaign messages that will help citizens and users be better prepared for how AI will present themselves online. You should consider some best practice approaches for how it can be used.
I’ve spent some time interacting with data scientists lately, and I disagree that generative AI will perform human-level intelligent tasks. There are limits to how “intelligent” this technology can be. Adult supervision required. That’s why companies make progress by actually limiting their datasets and focusing on where AI excels.
With that in mind, do you think Derek’s call to action is overly harsh? because it contains information. It is within AI’s current capabilities and we need to act.
Recommended by Diginomica – This Week’s Diginomika Top Stories
Vendor analysis, Diginomica style. My top three choices from vendor coverage are:
- FinancialForce talks about the need for connected processes in service organizations – Phil details FinancialForce’s quarterly release updates.But what’s interesting is that most of the service organizations FinancialForce has spoken to about customer success products have yet to see this feature as working in concert with their service delivery teams.“
- State of SAP – DSAG explains its position on hybrid cloud and S/4HANA feature parity. This is a pivotal time for SAP customers, and his DSAG, the German-speaking SAP user group, has a lot to say. Has his DSAG position on cloud changed and why on-premises feature parity is the hottest topic?Dive in our latest interview with DSAG.
- Dassault uses generative AI to create big context. George has another enterprise his angle on generative AI. This includes his IoT.Data from physical objects, supply chain partners, and third-party data providers is a bit fuzzy. In such cases, AI-powered data translators can go a long way to contextualizing information appropriately to the problem at hand.“
More vendor selection without quotes:
John’s Grab Bag – Gary wrote an inspirational use case, “How StepUp.One provides global online opportunities for refugee-only recruits.” Phil shares highlights from the Trailhead event on how Salesforce CMO Sarah Franklin and his Trailblazers build new careers in tech. Finally, Derek highlights the potential for AI surveillance amid calls from the UK data protection regulator to broaden the scope of the “AI fairness” principle to include development as well as use of AI. We are talking about expansion.
Best of Enterprise Web
my top 7
- Workday AI and ML Innovation Summit: Chasing the Eye of the AI Storm – In-depth review by Hyoun Park of Amalgam Insight. Includes insights on generative AI that applies beyond Workday.
- The “sell to enterprise” or “sell to IT” debate is a thing of the past – Gartner’s Hank Burns has it done: “The new reality is that in smart organizations that see technology as a source of advantage, business and IT are working together to improve business outcomes. In the old days, they fought for power and control. This is not the majority, but for providers, the ‘get around IT’ strategy can increasingly reduce profits.“
- Why an Independent ERP Software Consultant is Critical to Successful Digital Transformation – Well, Eric Kimberling sneaks in a third-stage consulting plugin here, but the fact remains: project bias is real. Yes, independent advisors are important: “Project bias continues to be an issue even after software selection… During implementation, project bias undermines the process by hiding project risks and problems to maintain revenue streams and software footprint. may turn into.”
- Multiplier Effect: How B2B Winners Grow – The consumerization of B2B has definitely begun.Nearly 70% of decision makers are prepared to spend up to $500,000 on a single ecommerce transaction.”
- The Great Flowering: Why OpenAI is the new AWS and the new Kingmakers are still important – I was waiting to hear James Governor’s take on the impact of OpenAI on developers.
- “Overhired” hustlers abuse ChatGPT to get more full-time jobs Important. Just to point out, these seem like highly skilled individuals to me who could do more, but ChatGPT can’t really handle their entire job, They can’t even close. Still, the drudgery they’ve cut back on is an early indicator of how work will change.
- ChatGPT is creating fake Guardian articles. Here’s our response – another interesting angle of ChatGPT went wrong. But The Guardian’s statement on the potential future use of generative AI was even more interesting (albeit still in its early stages).
capricious
That’s why its debut to Netflix’s live streaming didn’t look good, sparking all kinds of Love is Blind pannery and social media satire. My favorites came from Blockbuster:
I know you’ve had a rough round when a blockbuster lands with a right hook.This is one use case of generative AI that I’m not a fan of.
What the world needs, more mundane, contentless press releases. Fortune cookies on the other hand? spot:
Perhaps the earning potential is a bit limited, but… #ensw For good or bad hits and misses, let us know in the comments: Clive (Almost) always. Most enterprise hits and misses articles are selected from my handpicked articles @jonerp newsfeed.
