DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
2. A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant’s submission filed on 03/02/2026 has been entered.
3. Claims 1, 11, 12, 19 and 20 have been amended; and therefore, claims 1-20 are currently pending in this application.
Claim Rejections - 35 USC § 101
4. Non-Statutory (Directed to a Judicial Exception without an Inventive Concept/Significantly More)
35 U.S.C.101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
● Claims 1-20 are rejected under 35 U.S.C.101 because the claimed invention is directed to an abstract idea without significantly more.
(Step 1)
The current claims fall within one of the four statutory categories of invention (MPEP 2106.03).
(Step 2A) [Wingdings font/0xE0] Prong One:
The claim(s) recite a judicial exception, namely an abstract idea, as shown below:
— Considering each of claims 1, 11 and 20 as representative claims, the following claimed limitations recite an abstract idea:
Claim 1:
a prompt assignment comprising instructions for teaching a user how to construct prompts [for] conversational exchanges; a user-created prompt in response to the prompt assignment; receive a reply to the user-created prompt; [draft] a prompt insight evaluating the effectiveness of the user-created prompt in eliciting a desired response, wherein the evaluation [is] based on one or more interaction metrics comprising a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt; and [present] the reply together with the prompt insight in the context of the prompt assignment.
Claim 11 and 20:
a prompt assignment comprising instruction for teaching a user how to construct prompts [for] conversational exchanges; a first user-created prompt based on the prompt assignment; receive a first reply to the first user-created prompt; [draft] a prompt insight evaluating the effectiveness of the first user-created prompt in eliciting a desired response, wherein the evaluation [is] based on one or more interaction metrics comprising a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt; and provide the first reply and the prompt insight.
Thus, the limitations identified above recite an abstract idea since the limitations correspond to certain methods of organizing human activity, and/or mental processes, which are part of the enumerated groupings of abstract ideas identified according to the current eligibility standard (see MPEP 2106.04(a)).
For instance, the current claims correspond to managing personal behavior (e.g., teaching), wherein a user is presented with a prompt assignment, which comprises instructions for teaching a user how to construct prompt for conversational exchanges; and thus, based on a user-created prompt, which is received from the user in response to the prompt assignment, the user is presented with pertinent information; namely: (i) a reply to the user-created prompt, and (ii) a prompt insight evaluating the effectiveness of the user-created prompt in eliciting a desired response, wherein the evaluation is based on interaction metrics comprising a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt, etc.
Similarly, given the fact that the core of the claimed process can be performed in the human mind (and/or using a pen and paper), including the limitations that recite the process of: presenting a prompt assignment comprising instructions for teaching a user how to construct prompts; generating a reply to the user-created prompt; generating a prompt insight evaluating the effectiveness of the user-created prompt in eliciting a desired response, wherein the evaluation is based on one or more interaction metrics comprising a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt, etc., the claims also overlap with the group mental processes—i.e., observation, evaluation, judgment, opinion, etc.
(Step 2A) [Wingdings font/0xE0] Prong Two:
The claim(s) recite additional element(s), wherein a computer device/system with basic components (e.g., a processor, a storage media, etc.), which executes an algorithm—such as, a generative artificial intelligence model, is utilized to facilitate the recited functions/steps regarding: displaying information to a user (e.g., “present, via a learning platform, a prompt assignment comprising instructions for teaching a user how to construct prompts for submission to a generative artificial intelligence (AI) model, wherein the generative AI model supports conversational exchanges with the user”); collecting a response(s)/input(s) from the user (e.g., “receive, via a user interface of the learning platform, user input comprising a user-created prompt in response to the prompt assignment”); analyzing the collected response/input using an algorithm (e.g., “submit the user-created prompt to the generative AI model”); and generate/present pertinent one or more results based on the analysis above (“receive, from the generative AI model, a reply to the user-created prompt; generate, by an insights service integrated with the learning platform, a prompt insight evaluating the effectiveness of the user-created prompt in eliciting a desired response . . . based on one or more model-interaction metrics comprising a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt; and display the reply together with the prompt insight in the context of the prompt assignment”), etc.
However, the claimed additional element(s) fail to integrate the abstract idea into a practical application since the additional element(s) are utilized merely as a tool to facilitate the abstract idea. Thus, when each claim is considered as a whole, the additional element(s) fail to integrate the abstract idea into a practical application since they fail to impose meaningful limits on practicing the abstract idea. Although the claims utilize an algorithm—namely, a generative artificial intelligence (AI) model—to generate pertinent information (e.g., a reply and a prompt insight) based on the analysis of input data collected from the user (e.g., input in the form of a user-created prompt), neither the current claims nor the original disclosure signifies any technological improvement in this regard. Instead, each of the current claims, including the original disclosure, is utilizing the existing computer/network technology—merely as a tool—to facilitate the presentation of pertinent information to the user, based on the analysis of input data collected from the user, etc.
Accordingly, when each of the claims is considered as a whole, none of the claims provides an improvement over the relevant existing technology.
The observations above confirm that the claims are indeed directed to an abstract idea.
(Step 2B)
Accordingly, when the claim(s) is considered as a whole (i.e., considering all claim elements both individually and in combination), the claimed additional elements do not provide meaningful limitations to transform the abstract idea into a patent-eligible application of the abstract idea such that the claim(s) amounts to “significantly more” than the abstract idea itself (also see MPEP 2106). The claimed additional elements are directed to conventional computer elements, which are serving merely to perform conventional computer functions.
Accordingly, none of the current claims, when considered as a whole, recites an element—or a combination of elements—directed to an inventive concept.
It is also worth to note that the utilization of the conventional computer/network technology to facilitate the presentation of pertinent information to a user, including facilitating a teaching/learning process, wherein the user is presented with one or more pertinent feedback based on the analysis of the user’s response(s) to one or more assignments, etc., is directed to a well-understood, routine, conventional activity in the art (e.g., US 2016/0343272; US 2017/0091312; US 2012/0329029; US 2009/0198488; US 2012/0034591, etc.).
Note also that the use of a generative artificial intelligence, as an interactive tool, to facilitate human-machine interaction, including engaging a user with a more realistic human-like natural conversations, etc., is directed to a well-understood, routine, conventional activity in the art (e.g., see US 2018/0020093; US 2018/0376002; also see “Deep Reinforcement Learning For Dialog Generation”, the Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1192–1202, Austin, Texas, November 1-5, 2016. ©2016 Association for Computational Linguistics.
The observations above confirm that the current claimed invention fails to amount to “significantly more” than an abstract idea.
It is worth noting that the above analysis already encompasses each of the current dependent claims (i.e., claims 2-10 and 12-19). Particularly, each of the dependent claims also fails to amount to “significantly more” than the abstract idea since each dependent claim is directed to a further abstract idea, and/or a further conventional computer element(s) utilized to facilitate the abstract idea.
Accordingly, the findings above demonstrate that none of the claims implements an element—or a combination of elements—directed to an inventive concept (e.g., none of the current claims is reciting an element—or a combination of elements—that provides a technological improvement over the existing/conventional technology).
► Applicant’s arguments directed to section §101 have been fully considered (the arguments filed on 03/02/2026). However, the arguments are not persuasive at least for the following reasons:
Firstly, regarding Prong One of Step 2A, Applicant is asserting that “claim 1 is not directed to a judicial exception because it does not recite a mental process, a method of organizing human activity, or a mathematical concept. . . Instead, claim 1 recites a machine-implemented control process that measures and acts upon computer-native signals generated by a generative AI system” (emphasis added).
However, Applicant appears to fail to properly apply the inquiry established per Prong One of Step 2A. In particular, unlike Applicant’s assertion above, Prong One of Step 2A does not consider any of the claimed computer elements. Instead, Prong One is concerned with identifying merely the limitations that recite a judicial exception (e.g., an abstract idea); also see MPEP 2106.07(a), emphasis added).
For Step 2A Prong One, the rejection should identify the judicial exception by referring to what is recited (i.e., set forth or described) in the claim and explain why it is considered an exception. For example, if the claim is directed to an abstract idea, the rejection should identify the abstract idea as it is recited (i.e., set forth or described) in the claim and explain why it is an abstract idea.
In contrast, Applicant is referring to the computer elements—namely, the alleged machine-implemented control process, which supposedly “measures and acts upon computer-native signals generated by a generative AI system”, in an attempt to challenge the Office’s findings presented under Prong One of Step 2A. Consequently, Applicant’s arguments are not even relevant to challenge—much less negate—the Office’s findings under Prong One of Step 2A.
In addition, while attempting to compare current claim 1 with some of the examples in the USPTO guidance, Applicant asserts, “[t]his is analogous to Claim 2 of Example 37 in the USPTO's Subject Matter Eligibility Guidance . . . Likewise, amended claim 1 requires the insights service to evaluate the effectiveness of a user-created prompt based on model-interaction metrics, including a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt. These metrics are generated by and stored within a generative AI system and its interaction logs, and therefore cannot be observed, tracked, or computed by a human being using pen and paper or mental observation . . . as in Example 37, where icon rearrangement was driven by measured memory usage, the present claim drives prompt evaluation based on measured AI-system behavior. A human user may see an Al's reply but cannot directly track or compute the number of conversational turns, quantify response length across interactions, or determine how many prompt revisions were generated and submitted by the system. Those are computer-internal telemetry values that exist only because the generative AI model and learning platform record and process them” (emphasis added).
However, unlike Applicant’s assertions above, none of the current claims is even remotely analogous to Claim 2 of Example 37. In particular, unlike the current claims, Claim 2 of Example 37 does not recite any judicial exception (e.g., an abstract idea). Instead, Example 37 is referring to tracking computer-app parameters, as opposed to user’s prompt attributes. In particular, Claim 2 of Example 37 is tracking the amount of memory-space allocated to each computer-app of a plurality of computer-apps. This fact itself signifies that Claim 2 of Example 37 necessarily requires a computer device since computer-apps are normally executed via a processor of a computer. In contrast, the current claims are concerned with evaluating the effectiveness of a prompt that a user is making; namely, evaluating—using one or more prompt metrics—how effective the user’s prompt (e.g. the user’s question or remark) is to elicit a desired response. Of course, once the effectiveness of the prompt is evaluated, the user is presented with relevant information (e.g., a pertinent reply, the prompt insight, etc.). Accordingly, unlike the case of Claim 2 of Example 37, one does not even need a computer, much less a computer that executes an AI algorithm, to achieve the above objective. This is again because a human—such as a language instructor—can easily perform the above process mentally and/or using a pen and paper. The fact above already demonstrates the reason why none of the current clams is even remotely analogous to Claim 2 of Example 37.
Moreover, unlike the current claims, Claim 2 of Example 37 is automatically configuring computer features. For instance, based on the evaluation of each of the plurality of computer-app parameters (note that a computer-app parameter represents the amount of memory space allocated to the computer app), which the computer is tracking in its memory over a predetermined period of time, the computer automatically configures the features of the computer—namely, the GUI of the computer—by
automatically rearranging a plurality of computer-app icons that are linked to the computer apps (i.e., it automatically moves the most used icons to a position on the GUI closest to the start icon based on the determined amount of use). In contrast, none of the current claims performs any operation to enhance any features of the computer system. Instead, the current claims, including the original disclosure, are concerned with providing the user with relevant assistance based on the evaluation of one or more prompts that the user is making. Thus, the fact above is yet another reason why Claim 2 of Example 37 is not even remotely relevant to any of the current claims.
Applicant also appears to improperly apply the eligibility test established per Prong One, which one must apply when evaluating a mental process. For instance, unlike Applicant’s theory, the test regarding a mental process has nothing to do with evaluating whether a human—e.g., the instructor—is able to see the reply that an AI system is generating and/or the various computations that the AI system is supposedly performing. This is because the test does not even consider the existence of the AI system itself, much less the operations it is performing in order to provide prompting assistance to the user. Instead, the test regarding a mental process is concerned merely with determining whether a human—e.g., an instructor—can provide such assistance to the user. In this regard, while observing one or more prompts that the user is making when attempting to draft a proper prompt that elicits a desired response, the instructor evaluates mentally—and/or using a pen and paper—the effectiveness of the user’s prompt. In particular, the instructor can use one or more metrics for evaluating the effectiveness the user’s prompt; and such factors include: the number of conversational turns, the length of the response, the number of revisions that the instructor made to the user’s prompt, etc. Of course, once completing the above evaluation, the instructor presents the user—verbally and/or via pen and paper—with relevant feedback/information (e.g., prompt insight in the context of the prompt assignment, etc.).
The observation above demonstrates how the Prong One test is correctly applied when analyzing whether the claim is reciting a mental process. In particular, none of the computer elements, including the AI system, is considered since none of the claimed computer elements is part of the abstract idea. Instead, the computer elements are part of the additional elements. Consequently, Applicant’s arguments are not persuasive.
Similarly, regarding certain methods of organizing human activity, Applicant is asserting that “[w]hile the claim involves a ‘prompt assignment,’ the determination of prompt effectiveness is not based on pedagogical judgment or human evaluation, but on quantitative, system-level interaction data produced by the generative AI model itself. This is no different in principle from Example 37, where icon placement was not based on user preference or subjective judgment, but on objective memory-usage data tracked by a processor . . . under the same reasoning applied by the USPTO in Example 37, amended claim 1 does not recite a judicial exception at all. It is directed to a technical process for measuring and responding to generative-AI interaction data . . . satisfies Step 2A, Prong One” (emphasis added).
However, here also Applicant appears to fail to properly apply the eligibility test established per Prong One of Step 2A. In particular, Applicant is once again relying on the computer elements in order to challenge the Office’s findings regarding the group certain methods of organizing human activity. In contrast, as already pointed out above, Prong One does not consider any of the computer elements; such as, Applicant’s alleged “system-level interaction data produced by the generative AI model” and/or the alleged “objective memory-usage data tracked by a processor”, etc. Instead, Prong One is concerned merely with determining whether the claim is reciting an abstract idea; and if so, identifying merely the limitations that recite the abstract idea. Thus, considering current claim 1, the claim not only receives a user-created prompt, which the user is providing in response to a prompt assignment, but also provides the user with relevant feedback, which includes a reply to the prompt and a prompt insight, based on the evaluation of the effectiveness of the user’s prompt—i.e., how effective the user’s prompt is to elicit a desired response; and wherein such evaluation is based on one or more metrics (e.g., the number of conversational turns, the length of the response, the number of revisions made to the user’s prompt).
Thus, even basic common sense dictates that such process of providing the user with relevant information regarding the prompt/question that the user has constructed; such as, providing the user with a reply and/or suggested revisions to the prompt that the user has constructed, is indeed managing personal behavior (e.g., teaching), per the group, certain methods of organizing human activity. Consequently, Applicant’s attempt to challenge the Office’s finding under Prong One, while incorrectly conflating the computer elements with part of the abstract idea, is certainly not persuasive.
Note also that regarding Claim 2 of Example 37, the discussion presented above already demonstrates the reason why none of the current claims is even remotely analogous to Claim 2 of Example 37. Accordingly, Applicant’s attempt to challenge the Office’s finding under Prong One, while repetitively relying on such non-analogous example, is once again not persuasive.
Secondly, regarding Prong Two of Step 2A, Applicant asserts, “[t]he amended claim . . . recites a specific technological solution to a recognized technical problem arising from modern generative AI systems . . . a core technical problem is that unskilled users frequently submit ineffective prompts to generative AI models, which causes excessive conversational churn—multiple back-and-forth exchanges, repeated prompt revisions, and unnecessarily long responses. This results in substantial and unnecessary consumption of computing resources . . . this inefficiency produces a measurable, system-level impact on computing infrastructure and energy consumption . . . claim 1 directly addresses this technical problem by implementing a closed-loop, machine-driven prompt optimization system. The learning platform and insights service do not merely display feedback to a user; they measure how the generative AI model actually behaves in response to a prompt using model-interaction metrics . . . produces a concrete technological effect: fewer conversational turns, fewer prompt revisions, and shorter, more targeted AI responses, which in turn reduces the number of model executions, reduces token processing, and lowers the computational and energy cost of operating the generative AI system. The claimed invention therefore improves the functioning and efficiency of the generative AI system and the underlying computing infrastructure . . . applies any alleged abstract idea in a specific, technical way that improves how generative AI systems operate at scale” (emphasis added).
However, Applicant appears to be mistaking the assistance, which the computer system is providing to users, for an alleged technological improvement. In contrast, the currently claimed (and the originally disclosed) computer system does not implement any technological feature—or any combination of technological features—that provides a technological improvement over the relevant existing technology. In particular, the use of one or more AI models to analyze collected information (e.g., one more queries from one or more users); and subsequently providing each of the users with relevant information (e.g., one or more responses to a user’s query, one or more suggestions to the user’s query, etc.), is already part of the existing computer/network technology. Thus, Applicant’s claimed (and disclosed) system/method is still utilizing the existing computer/network technology—merely as a tool—to facilitate an abstract idea; such as, proving a user with pertinent information based on the analysis of collected information from the user, etc. (see any of the current claims). Consequently, Applicant’s alleged technological improvement is not persuasive.
In addition, besides incorrectly relying on the behavior of users to substantiate the alleged technological improvement, Applicant is also incorrectly considering the inherent attributes of the existing computer/network technology as the alleged technological improvement. For instance, unlike Applicant’s theory, the behavior of users—e.g., unskilled users who frequently submit ineffective prompts—has nothing to do with technology, much less a technological improvement. This is because, regardless of the number of skilled users who are submitting effective prompts, and/or regardless the number of unskilled users who are submitting ineffective prompts, the system is simply performing existing functions—i.e., collecting information, analyzing the information, and generating one or more relevant results. Thus, the computer is performing the same routine above, regardless of whether it is interacting with (a) skilled users only, (b) unskilled users only, or (c) a combination thereof. Of course, the content of the prompt (i.e., the topic) may vary from one user to another. In fact, a prompt that a first skilled user submits may be very different from a prompt that a second skilled user is submitting, let alone a prompt that an unskilled user is submitting. However, such different prompts that different users are submitting, and/or the different responses that the system is providing to the different users, etc., has once again nothing to do with technology, much less a technological improvement. Moreover, whether the user is a skilled user or an unskilled user, the user may interact with the system continuously or intermittently, regardless of whether the user is constructing an effective prompt or an ineffective prompt. In particular, the system is merely performing existing computer functions (i.e., collecting user’s input, analyzing the input using an algorithm, and generating relevant feedback to the user) regardless of whether the content/topic of the user’s input is assumed to be an effective—or an ineffective—prompt. Consequently, neither the technical problems that Applicant is alleging (e.g., “unnecessary consumption of computing resources”, “repeated execution of large language models”, “repeated token processing”, “increased server-side and client-side energy usage”), nor the technical solutions that Applicant is alleging (e.g., the so-called “closed-loop, machine-driven prompt optimization system” and/or the so-called “a concrete technological effect”, which supposedly “reduces the number of model executions”, “reduces token processing”, “lowers the computational and energy cost of operating the generative AI system”, “improves the functioning and efficiency of the generative AI system and the underlying computing infrastructure”, etc.), is even remotely relevant to demonstrate a technological improvement. Thus, Applicant’s arguments are not persuasive.
Moreover, given the features of the current claims and the description in the original disclosure, the claimed—and the disclosed—system is directed merely to the existing computer/network technology. In particular, as already pointed out above, it is part of the existing computer/network technology to implement one or more existing algorithms, including one or more generative AI models, in order to provide users with pertinent information (e.g., a response to the user’s question; an alternative question to confirm the user’s intent; a recommendation, etc.), based on the analysis of input data collected from users (e.g., a constructed question received from a user; a selection of a choice received from a user , etc.). Although a reference is not necessarily required to demonstrate the above fact, one or more of the references cited as part of the Step 2B analysis already confirm that the use of one or more AI models, which interact with users in a human-like natural conversations (e.g., receiving a request, providing a response and/or a suggestion, learning from the responses provided to the user, etc.), is already part of the existing computer/network technology. Accordingly, even if one considers the various technical solutions, which Applicant is alleging above (i.e., the alleged technical solutions that supposedly address the technical problems that Applicant is alleging), Applicant is effectively relying on the inherent features of the exiting computer/network technology. This is again because it is part of the existing computer/network technology to implement one or more generative artificial intelligence algorithms, including those with closed-loop models, that engage a user with a more realistic human-like natural conversations, etc. In fact, the implementation of an interactive closed-loop AI dates back to the 1970. For instance, SHRDLU, which is considered to be one of the earliest closed-loop AI models (developed by Terry Winograd at MIT between 1968 and 1970), not only receives a natural language command from the user, but also (i) considers previous interactions in order to resolve ambiguous terms, and (ii) uses recursive reasoning (e.g., tracing its logic back through its loop in order to provide the user with a proper response regarding the user’s current question, etc.).
The observations above confirm that Applicant is attempting to substantiate an alleged technological improvement while relying on the features and/or the benefits of the existing computer/network technology. Consequently, Applicant’s conclusory assertion, “the amended claim applies any alleged abstract idea in a specific, technical way that improves how generative AI systems operate at scale . . . it controls and optimizes AI-system behavior to reduce computational waste and energy consumption”, is once again not persuasive.
Secondly, regarding Step 2B, Applicant is asserting, “claim 1 amounts to significantly more than any alleged judicial exception because it recites a specific, machine-implemented architecture for controlling and optimizing generative AI system behavior . . . it defines a coordinated technical framework in which a learning platform, a generative AI model, and an insights service cooperate to measure model-interaction metrics and use those measurements to improve how the AI system operates” (emphasis added).
However, except for the labeling, “specific machine-implemented architecture”, Applicant fails to identify a feature (if any)—or a combination of features (if any)—that supposedly amount to “significantly more” than an abstract idea. It is worth noting that a claimed system/method is considered to be “significantly more” than an abstract idea if it is directed to the non-generic and non-conventional arrangement of the claimed additional elements. Of course, a technological improvement also demonstrates an inventive concept that amounts to “significantly more” than an abstract idea; also see MPEP 2106.05(a), (emphasis added),
While improvements were evaluated in Alice Corp. as relevant to the search for an inventive concept (Step 2B), several decisions of the Federal Circuit have also evaluated this consideration when determining whether a claim was directed to an abstract idea (Step 2A). See, e.g., Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1335-36, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016); McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1314-16, 120 USPQ2d 1091, 1102-03 (Fed. Cir. 2016); Visual Memory, LLC v. NVIDIA Corp., 867 F.3d 1253, 1259-60, 123 USPQ2d 1712, 1717 (Fed. Cir. 2017). Thus, an examiner should evaluate whether a claim contains an improvement to the functioning of a computer or to any other technology or technical field at Step 2A Prong Two and Step 2B, as well as when considering whether the claim has such self-evident eligibility that it qualifies for the streamlined analysis.
In contrast, Applicant is attempting to portray the alleged inventive concept while labeling the features of the conventional computer/network technology as the alleged “specific, machine-implemented architecture”, which supposedly controls and optimizes generative AI system behavior. Of course, Applicant has also attempted to label the conventional computer/network technology as the alleged “coordinated technical framework”, which supposedly cooperates a learning platform, a generative AI model, and an insights service to allegedly “measure model-interaction metrics and use those measurements to improve how the AI system operates”, etc. Thus, none of Applicant’s conclusory assertions are persuasive. This is again because Applicant fails to show a technological feature (if any)—or a combination of technological features (if any)—that is considered to be an advance over the existing computer/network technology. Instead, Applicant is repeatedly relying on the features of the conventional technology to substantiate the alleged technological improvement. For instance, per the conventional computer/network technology, AI models—including generative AI models—normally learn and update their parameters continuously, based on collecting and analyzing new/updated information; and such inherent process helps the models to improve the accuracy of the results that they are generating. Consequently, Applicant’s attempt to portray a technological improvement, while simply emphasizing such existing features of the conventional computer/network technology, is once again not persuasive.
Furthermore, while mistaking part of the abstract idea for an advanced technological feature, Applicant incorrectly asserts that “[t]he amended claims recite a telemetry-driven control mechanism in which interaction metrics—such as conversational turns, response length, and prompt revision counts—are measured and used to evaluate prompt effectiveness and drive AI interaction efficiency. The Office Action does not establish that this architecture, including the measurement and use of AI runtime interaction metrics for system-level optimization, was well-understood, routine, or conventional . . . claim therefore recites an inventive concept: a system that improves generative AI performance by reducing conversational churn, unnecessary model executions, and computing resource consumption through machine-measured prompt optimization. This is not a generic use of a computer to teach or display information; it is a technical improvement to the operation of a generative AI system itself” (emphasis added).
However, when determining whether a claim is directed to a well-understood, routine, conventional activity (hereinafter WRCA) in the art, such test is not referring to the abstract idea that the claim is reciting, regardless of whether the abstract idea is a
new abstract idea or an old one. Instead, the WRCA test—per Step 2B—is referring to the technology, but not necessarily to the abstract idea. In the instant case, part of the recited abstract idea, which requires the consideration the so-called “interaction metrics” that includes “conversational turns, response length, and prompt revision counts”, may be new data. However, this does not necessarily imply that the technology, which the claimed—and the disclosed—system/method is utilizing to facilitate the abstract idea above, is an advanced technology. This is once again because a claim for a new abstract idea is still an abstract idea; see Synopsys, Inc. v. Mentor Graphics Corp., 839 F.3d 1138, 1151 (Fed. Cir. 2016). Moreover, as repeatedly pointed out above, AI models—including generative AI models—are already part of the conventional computer/network technology; and these models continuously learn and update their parameters, based on collecting and analyzing new/updated information, in order to improve the results that they are generating. In contrast, Applicant is once again relying on such existing features of the conventional computer/network technology to portray the alleged technological improvement. Consequently, none of Applicant’s conclusory assertions regarding the alleged inventive concept—such as, the alleged “system that improves generative AI performance by reducing conversational churn, unnecessary model executions, and computing resource consumption through machine-measured prompt optimization”, and/or the alleged “technical improvement to the operation of a generative AI system itself”, etc., is even relevant to challenge—much less negate—the Office’s findings presented under Step 2B.
It is also noted that the original disclosure does not even support Applicant’s theory. For instance, per the original disclosure, the interaction metrics that the system is gathering (e.g., text of the prompt that the user has created, a revision made to the prompt, prompts submitted, dwell time over replies, etc.) are presented to a reviewer, so that the reviewer evaluates the performance of the user—such as, a teacher evaluating the user’s ability to create an effective prompt, etc. (e.g., see [0038] to [0050] of the specification). Accordingly, unlike Applicant’s assertion, the alleged model-interaction metrics are not even used to optimize or measure how the alleged generative AI model is behaving. Nevertheless, regardless of Applicant’s theory, the findings presented above already demonstrate the reasons why neither the current claims nor the original disclosure contemplates any technological improvement over the existing (conventual) computer/network technology.
Thus, at least for the reasons discussed above, the Office concludes that none of the current claims—when considered as a whole—implements an inventive concept that amounts to “significantly more” than an abstract idea.
Claim Rejections - 35 USC § 112
5. The following is a quotation of the first paragraph of 35 U.S.C.112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C.112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
● Claims 1-20 are rejected under 35 U.S.C.112(a) or 35 U.S.C.112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for pre-AIA the inventor(s), at the time the application was filed, had possession of the claimed invention.
Each of claims 1, 11 and 20 recites, “the insights service evaluates the effectiveness of the first user-created prompt based on one or more model-interaction metrics comprising a number of conversational turns, a response length, or a number of prompt revisions associated with the user-created prompt” (emphasis added).
However, the original disclosure does not have written description regarding such process of evaluating the effectiveness of the user’s prompt based on (i) a number of conversational turns, (ii) a response length and (iii) a number of prompt revisions associated with the user’s prompt. Instead, the original disclosure is describing the process of observing the prompting activities of the user; and wherein such observation includes observing: (a) the text of the prompt that the user created, (b) a revision made to the prompt subsequent to a suggestion made to the user, (c) the prompts submitted to the foundation model service, (d) replies that the foundation model service presented in response to prompts, (e) actions of users with respect to replies—such as, clicking links presented as part of replies, dwell time over replies (see [0401] of the specification).
The specification further describes that: (a) prompting activities may be organized into conversations between the user, the learning platform and the foundation model ([0043]), (b) the characteristics of observed user’s actions include: a dwell time over replies, a frequency of using a stop-replying feature with respect to replies and a frequency of click-throughs with respect to content of the replies ([0041]); (c) insights into the prompting are identified, which includes statistics—such as, counts related to the classification categories ([0045]) and also how many traditional Internet searches turned into chat conversations with a foundation model service ([0046]).
However, there appears to be no specific description regarding the process of evaluating the effeteness of the user’s prompt based on the three specific parameters currently claimed; namely, (1) the number of conversational turns, (2) the response length, and (3) the number of prompt revisions associated with the user-created prompt.
Prior Art
6. Considering each of claims 1, 11 and 20 as a whole (including the respective dependent claims), the prior art does not teach or suggest the current claims (regarding the state of the prior art, see the office action dated 07/30/2025).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRUK A GEBREMICHAEL whose telephone number is (571) 270-3079. The examiner can normally be reached on 7:00AM-3:00PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, DAVID LEWIS can be reached on (571) 272-7673. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/BRUK A GEBREMICHAEL/Primary Examiner, Art Unit 3715
Prosecution Insights