Thousands of AI agents later, who even remembers what they do?
- Reference: 1732188553
- News link: https://www.theregister.co.uk/2024/11/21/gartner_agentic_ai/
- Source link:
The collision of muddied management thinking and much-hyped autonomous agents will be interesting to watch play out.
In Gartner's view, "agentic AI" is about "goal-driven software entities that have been granted rights by the organization to act on its behalf to autonomously make decisions and take action."
[1]
They differ from robotic process automation, which stitches together enterprise applications with trained bots, as agents do not need explicit inputs and their outputs are not predetermined. The analyst company notes that AI agents have become the flavor of the month with vendors.
[2]
[3]
Notable examples include Salesforce, with the vendor's ebullient CEO, Marc Benioff, [4]boasting to investors that by releasing a billion agents by 2026, Salesforce could capture a "very high margin opportunity."
In its latest paper (available only to clients, although there is an upcoming open to all [5]webinar ), Gartner weighs up the pros and cons. Perhaps it's wisest to start with the bad news. "The danger exists of repeating the robotic process automation problem: organizations created thousands of bots, but nobody now remembers what those bots do or why they were built," it says. In addition, organizations might also build and deploy their own low-code agentic AI inside the IT stack, "which may not meet your security or quality standards."
[6]
"Agentic AI will make decisions based on its analysis of your organization's data, making plans based on that analysis. From there, it'll act on those plans. This will be dangerous unless you invest in the skills, practices and technologies to deliver trustworthy AI agents. Your organization's data may be of poor quality, further increasing the risk. As well as creating risk, poor data quality and architecture will also inhibit agentic AI's development."
Gartner also warned that AI agents could alienate customers if the experience is poorly designed.
Gartner says organizations should create "customer journey maps to design the ideal customer experience and define guardrails before handing over to AI agents for execution."
[7]AI PCs flood the market. Their makers hope someone wants them
[8]AI's power trip will leave energy grids begging for mercy by 2027
[9]Single-platform approach may fall short for AI data management
[10]Killer app for AI is still years away, says industry analyst
Whether it's a good idea or not, agentic AI is coming, Gartner says. It forecasts that a third of enterprise software will incorporate such agents while 20 percent of digital storefront interactions will be conducted by AI agents.
"Agentic AI will be incorporated into AI assistants and built into software, SaaS platforms, Internet-of-Things devices and robotics," Gartner said. "When AI assistants start planning, making decisions, and taking action for you, agentic AI will be there. It'll be everywhere, with the potential to extend collaborative work management platforms beyond task tracking into planning and executing tasks."
[11]
The benefit for organizations might be that – if employed properly – AI agents can "increase the number of tasks and workflows that can be automated."
"The potential that agentic AI has to constantly analyze the performance of personalized interactions surpasses human capabilities, ensuring more precise and effective customer engagement. Software developers are likely to be some of the first affected, as existing AI coding assistants gain maturity and AI agents provide the next set of incremental benefits," said the research paper, Top Strategic Technology Trends for 2025: Agentic AI.
Of course, risks can be managed for a reasonable cost if the potential benefits outweigh them. At this stage, though, we must ask ourselves whether any government or corporate IT department has ever been known to skimp on the hard part. ®
Get our [12]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2Zz9nMXKFsntpXb-3spzmmAAAAMM&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Zz9nMXKFsntpXb-3spzmmAAAAMM&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Zz9nMXKFsntpXb-3spzmmAAAAMM&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[4] https://www.theregister.com/2024/08/29/salesforce_pricing_per_ai_conversation/
[5] https://www.gartner.com/en/webinar/671130/1500131
[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44Zz9nMXKFsntpXb-3spzmmAAAAMM&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[7] https://www.theregister.com/2024/11/14/ai_pc_shipments/
[8] https://www.theregister.com/2024/11/13/datacenter_energy_consumption/
[9] https://www.theregister.com/2024/11/07/data_platform_vendors_ai/
[10] https://www.theregister.com/2024/11/07/it_spend_europe_2025/
[11] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33Zz9nMXKFsntpXb-3spzmmAAAAMM&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[12] https://whitepapers.theregister.com/
Re: Doh! Nuts!
Disaster averted. Good boy.
'Agentic'? Really? We're just making up words now I guess.
We've done it for as long as we've been talking. That's how words come to exist. In this case, however, it happened some time ago as DDG pulls up plenty of examples.
Our company has decided to fill Teams with bots. They can't be uninstalled, are named something cute that bears no relation to their function so who knows what they're for now, and all they seem to do randomly and suddenly nag me to sign in from time to time when I'm already signed into Teams. Thank goodness we are now all more productive.
That seems like somebody's being clever. Two birds with one stone: get Teams and AI banned together.
The room/desk-booking system at my company now features an AI "assistant". I asked it to tell me the airspeed velocity of an unladen swallow, but it did not understand the question. It was also unable to tell me the answer to Life, the Universe and Everything. Clearly, its training data did not include any of the classics.
Their outputs are not predetermined
That sounds as good a basis any to entrust an impenetrable black box to autonomously make decisions and take action.
It's all just Bullshit
AI seems good on initial glance because it's basically a pretty-good bullshit generator. But as soon as you ask questions about accuracy, truthfulness, and judgement you find AI is actually not very much use to anyone.
If you write software by copying patterns of code you've found on the internet, without understanding what they do or how they work, your code will be useless. That's what AI code writing agents do: they simply find something similar that appears to match what you've typed so far. It doesn't have any concept of "understanding" or "intelligence".
Re: It's all just Bullshit
I've yet to see an AI product that I would actually use.
I've never seen one that does something I couldn't do better with the same data, a tiny bit of code, and a slow processor, whereas they cost billions to train an inadequate statistical model.
It's glorified autocomplete (especially LLMs, which are literally this) but without the rigour and predictability. And I don't see any evidence that they actually "learn"... if they learned you wouldn't have to keep "retraining" the model... you could just expose it to the data, provide it with corrections and off it would go on the SAME MODEL adjusting as it encountered new information.
They reek of an over-complicated statistical model that - like all statistical AI models - require you to "train" undesirable behaviour out by overwhelming it statistically. If you trained it one 1,000,000 pieces of incorrect data, to train it to be correct you need to retrain with 1,000,001 pieces of data to the contrary, and so on. It's just working on the probability it finds in the dataset, nothing else.
And the nonsense about just using data phrases to CONTROL ITS RESPONSE (where the LLM creators provide it with huge hidden initial prompts that tell it what it should and should not answer, etc.) is the biggest lot of manure I've ever seen in my life. It's so easily overridden (again, statistically) and so un-rigorous that it's worthless.
Another few years and the fad with die and be consigned to the "oh, look, it can paint a picture" levels of app again, and we can get on with some real work on AI again. Like actually trying to solve the inference problem (which is what you're talking about) and not just continue building statistical models that we somehow hope will magically learn, turn intelligent and form AGI spontaneously.
Since the 60's the cries have been "if only we have more processing power", "if only we had more connectivity", "if only we had more training data", "if only we had more funds"... then AGI will just jump out of the ether like the soul of a person and somehow magically become intelligent.
Turns out, now that we're literally spending billions of dollars training models on billions of nodes, each with billions of instructions per second of processing, on the ENTIRE INTERNET of data, that just hasn't happened.
So maybe now is finally the time to just go away quietly and think about WHY that is and WHAT we actually need to do, rather than cross our fingers and hope Frankenstein's monster just jolts into life after cobbling some body parts together.
I've been saying that ever since I studied AI in the late 90's, and nothing has changed except that we now have LITERALLY what people were asking for... and it's made absolutely no difference to the actual intelligence of the system.
Re: It's all just Bullshit
That just reminded me. A while ago someone showed me about 10 lines of C written by an AI. Although I can't remember what it was for, I do remember that it worked, but I could thin it down to 3 or 4 lines.
Re: It's all just Bullshit
"it's basically a pretty-good bullshit generator. But as soon as you ask questions about accuracy, truthfulness, and judgement you find AI is actually not very much use to anyone."
But you repeat yourself.
The only bot worth looking at...
[1]https://inspirobot.me/
[1] https://inspirobot.me/
Corporate forgetfulness
> but nobody now remembers what those bots do or why they were built
Clearly they need to create a bot that asks all these bots to say what their purpose is.
"The collision of muddied management thinking and much-hyped autonomous agents will be interesting to watch play out."
The collision of muddied management thinking and much-hyped anything is SOP.
Doh!
Yet another reason I'm so glad I'm retired. Commiserations for you folks having to deal with this crap.