Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Applicant’s arguments, see p. 11, filed 06/30/2026, with respect to the 35 U.S.C. § 101 rejection of claims 1-20 have been fully considered and are persuasive. Considering the amendments filed 06/30/2026 structure the claims to identify what learning models are used and concrete details of how they are used together to accomplish certain tasks, which serves as evidence of improvements to the function of certain computers, the 35 U.S.C. § 101 rejection of claims 1-20 has been withdrawn.
Applicant’s arguments with respect to the 35 U.S.C. § 102 rejection of claims 1-15 and 17-20 under Rao have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 3-11, 13, 15, 19 and 21-22 are rejected under 35 U.S.C. 103 as being unpatentable over US 2023/0381664 A1 to Desgarennes et al.
Re claim 1, Desgarennes teaches: An apparatus, comprising: at least one processor system programmed with instructions to: execute a video game to render a single-player game instance;
[0020], Desgarennes is a system and method that uses machine learning models including language models, audio models, speech models and multimodal learning models to provide personalized agents/bots in game environments.
[0023], the personalized agents can act as a constant user companion as the user plays different games and plays different game modes.
[0024], a gaming device such as a mobile device or console gaming system executes a game locally or on a server such as a cloud server.
Fig. 1A, a cloud service 104 comprising game service 106 is in bidirectional communications with user device 102.
Regarding the limitation “…to render a single-player game”, [0037] of Degarennes discloses “any number of players can participate in a gaming session using system 150”
while the video game is executing, receive an audible prompt from a first player, the audible prompt indicating a task for an artificial intelligence (AI) model architecture
[0021]-[0022], user input may comprise spoken natural language inputs.
[0025], conversational “voice commands” are received from user device 102.
[0026], an example provided is a user speaking “cover the right side” while a first-person shooter game is being executed. This serves as a task for a model 118 to interpret it. [0064] is another example where a user asks “Does this object look like a house” via a speech or chat interface.
[0032], “one or more models 116 may process video, audio … received from the user … in order to interpret user commands and/or derive user intent based upon the user communications”
comprising a first model, a second model, and a third model to execute in support of the first player in the single-player game instance;
[0020] enumerates more than three models that are all usable in support of a first-player in a game instance. These include reinforcement learning models, foundation models, language models, computer vision models, speech models, video models, audio models, and multimodal machine learning models.
execute the first model, comprising a large language model (LLM), to make inferences of relevant in-game tasks to execute based on current game state data for the video game and the audible prompt
[0050] "... the personalized agent is operable to receive visual, audio, and textual game data that is available to a human user, process the received data using one or more machine learning models to determine a current game state, and determine appropriate actions to perform based upon the current game state."
[0062], received game data may be provided to “natural language understanding models … in order to process the current game state.”
execute the second model, comprising an audio generator, to present an audible output in a voice of a video game character from the video game itself based on an output from the first model:
[0050]-[0051], [0056], [0062], personalized agent actions are determined and used to control agents in the game space based on analyzed and determined game state.
And note that [0021] states that a generative multimodal learning model generates multimodal output that may comprise natural language output.
Regarding presenting generated audio output in a game character’s voice, [0022] states that model output can be multimodal and include "spoken or written language (which may also be referred to herein as "natural language output")". And the model outputs may take the form of a personalized non-player character (NPC) agent that can follow a player character through different games, see [0025]. [0039] notes that agents can be selected from a plurality of possible builds from an agent library and be further configured by the user.
It would have been obvious, if not inherent, that audio/video natural language speech output by a learning model of one of a plurality of personalized NPC agents operating from an agent library would have involved a voice of the character or else the NL output would not qualify as speech.
execute the third model, comprising a machine learning-based gameplay model trained through reinforcement learning, to issue control commands to a game engine for the Al model architecture to play the video game based on the output from the first model; and based on a trigger, present the audible output in the voice of the video game character while concurrently issuing the control commands to the game engine to cause the Al model architecture
[0035], gameplay machine learning models 128 receive data from the executed game ... and generate control signals to direct the gameplay of the agent within the game ... through reinforcement learning ... agents can be trained to play specific games ... personalized agent library 107 is operable to receive trained models .... and employ those models ... to control the instantiated agent(s) 108."
[0051], one or more ML models transmits control instructions to control gameplay of a personalized agent based on ML-model learned game state and determined agent actions.
[0056], control signals may be received to control gameplay of the personalized game agent within the game instance based on determined game state and determined, generated actions.
See additionally [0033], [0040], interpreted game states are used to generate instructions to control personalized agents actions. [0061]
to play the single-player game instance as a virtual teammate along with the first player.
[0004], [0023], [0035], [0036] describe the AI agent as a “constant gameplay companion” and “useful companion” who can assist the user through the play of a variety of different games. [0029] describes that an agent may suggest that a player changes from a front-line attacker to a support character in a game. All this functionality qualifies as a virtual teammate.
Re claims 13, 19, refer to the rejection of claim 1.
Re claims 3, [0026], [0049], [0060], the game may be a first-person shooter game.
Re claims 4-5, [0004], [0023], [0035], [0036] describe the AI agent as a “constant gameplay companion” and “useful companion” who can assist the user through the play of a variety of different games. [0029] describes that an agent may suggest that a player changes from a front-line attacker to a support character in a game. This functionality qualifies as being an ally/team member.
Re claims 6-11, 25 and 27, these claims recite 1) characteristics or attributes intended for a second character, but lacking any additional method steps required to provide them and 2) the content of data representing “tasks” that are usable “for a model” in parent claim 1. Parent claim 1 does not require these “tasks” to be executed in accordance with any particular software instructions but merely identifies they are capable of use for some model itself having unclaimed hardware and/or software. "[A]pparatus claims cover what a device is, not what a device does." Hewlett-Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464, 1469, 15 USPQ2d 1525, 1528 (Fed. Cir. 1990) (emphasis in original). A claim containing a "recitation with respect to the manner in which a claimed apparatus is intended to be employed does not differentiate the claimed apparatus from a prior art apparatus" if the prior art apparatus teaches all the structural limitations of the claim. Ex parte Masham, 2 USPQ2d 1647 (Bd. Pat. App. & Inter. 1987).
Additionally, these claims are seen as being directed to the particular payload or attributes of data (equivalent to claims to “data per se”), namely character attributes and types of tasks that are merely claimed functionally but not required to be used in any particular steps assigned to a programmed apparatus. As indicated in MPEP § 2106.03, data per se and software per se do not belong to a statutory category of invention. When considered as an additional element under the Alice analysis, then, matters of data per se cannot impart eligibility to judicial exceptions.
Restricting data in a claim to a particular type or content is also seen as filtering gathered data, which has been held in Bascom Global Internet v. AT&T Mobility LLC, 827 F.3d 1341, 1349, 119 USPQ2d 1236, to represent an abstract idea of the grouping of “certain methods of organizing human activity” because interacting with a database and selecting certain data by generic computers is not fundamentally different from human beings interacting with printed content. “[f]iltering software, apparently composed of filtering schemes and filtering elements, was well-known in the prior art” and “using ISP servers to filter content was well-known to practitioners.”
Re claim 15, outputs include images, video, and visual features in addition to natural language output.
Re claim 21, claims such as this directed to the appearance of computer graphics on a display device, wherein the appearance of graphics is solely intended to convey meaning to a human user and lacking any functional alteration or transformation of the device that would define it as a new apparatus, are not awarded patentable weight because they are considered nonfunctional descriptive material (printed matter). See MPEP 2111.05 Part III which explains that, for computer-readable media, where the claim is directed to conveying a message or meaning to a human reader and/or the computer-readable medium merely serves as a support for information or data, no functional relationship exists and the graphics are not of patentable significance.
Re claim 22, [0020] describes that agents can recognize text and [0022] describes that outputs from agents can be natural language output.
Re claim 23, [0039] notes that agents can be selected from a plurality of possible builds from an agent library and be further configured by the user.
Re claim 24, [0035] describes that gameplay machine learning models 128 can use reinforcement learning to develop an understanding of gameplay mechanics for specific games and genres of games.
Re claim 26, as noted in [0022], agent outputs may be in natural language format, and as noted in [0029], agent outputs can suggest a player modify his playstyle to achieve a certain objective.
Claim 16 is rejected under 35 U.S.C. 103 as being unpatentable over US 2023/0381664 A1 to Desgarennes et al. in view of US 20200269136 A1 to Gurumurthy et al.
Re claim 16, Although Desgarennes teaches substantially the same inventive concept including outputting multimodal suggestions to a game player generated using machine learning, Rao is silent as to whether haptic outputs can be generated.
Gurumurthy is an analogous prior art reference for gamer training using neural networks. Gurumurthy teaches, see [0012], [0019], that it was known in this art for “haptic feedback or guidance in near real time during game play” to be an alternative to or addition to “providing visual, audio” feedback or guidance as “the recommended next action for the player to take in the game.”
It would have been obvious to one having ordinary skill in the art before the effective filing date of the instant invention that Desgarennes’ multimodal machine learning gamer assistance could have outputted haptic (tactile) output instead of or in addition to visual and audio guidance as taught by Gurumurthy without causing any unexpected results. An expected advantage would be that the human body is highly sensitive to haptic feedback.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to STEVEN J HYLINSKI whose telephone number is (571)270-1995. The examiner can normally be reached Mon-Fri 10-530.
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, Dmitry Suhol can be reached at (571) 272-4430. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/STEVEN J HYLINSKI/Primary Examiner, Art Unit 3715
Prosecution Insights